BLAST is a high-performance serving engine for browser-augmented LLMs, designed to make deploying web-browsing AI easy, fast, and cost-manageable.
The goal with BLAST is to ultimately achieve google search level latencies for tasks that currently require a lot of typing and clicking around inside a browser. We're starting off with automatic parallelism, prefix caching, budgeting (memory and LLM cost), and an OpenAI-Compatible API but have a ton of ideas in the pipe!
Website & Docs: https://blastproject.org/ https://docs.blastproject.org/
MIT-Licensed Open-Source: https://github.com/stanford-mast/blast
Hope some folks here find this useful! Please let me know what you think in the comments or ping me on Discord.
\u2014 Caleb (PhD student @ Stanford CS)"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: Blast \u2013 Fast, multi-threaded serving engine for web browsing AI agents"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/stanford-mast/blast"}},"_tags":["story","author_calebhwin","story_43872761","show_hn"],"author":"calebhwin","children":[43872846,43872877,43873068,43873364,43873396,43873595,43873752,43873986,43874249,43874469,43874681,43875778,43877124,43879134,43880489],"created_at":"2025-05-02T17:42:28Z","created_at_i":1746207748,"num_comments":66,"objectID":"43872761","points":145,"story_id":43872761,"story_text":"Hi HN!
BLAST is a high-performance serving engine for browser-augmented LLMs, designed to make deploying web-browsing AI easy, fast, and cost-manageable.
The goal with BLAST is to ultimately achieve google search level latencies for tasks that currently require a lot of typing and clicking around inside a browser. We're starting off with automatic parallelism, prefix caching, budgeting (memory and LLM cost), and an OpenAI-Compatible API but have a ton of ideas in the pipe!
Website & Docs: https://blastproject.org/ https://docs.blastproject.org/
MIT-Licensed Open-Source: https://github.com/stanford-mast/blast
Hope some folks here find this useful! Please let me know what you think in the comments or ping me on Discord.
\u2014 Caleb (PhD student @ Stanford CS)","title":"Show HN: Blast \u2013 Fast, multi-threaded serving engine for web browsing AI agents","updated_at":"2026-05-13T11:27:19Z","url":"https://github.com/stanford-mast/blast"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"inssein"},"story_text":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"Hi All,
I\u2019ve been working on Aidmin for quite some time now, and finally released a beta version of it.
I\u2019ve focused most of my efforts on security, making sure that Aidmin can\u2019t negatively impact your database. I talk about that a little bit in the Security Overview (https://github.com/aidmin-io/docs/blob/main/security-overvie...).
I would love to know if Aidmin would be helpful at your workplace, and if not, why? Are there any features that are missing which would make it much more useful?
Thank you for checking out the project!"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: Aidmin \u2013 Web-based database GUI"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"https://aidmin.io"}},"_tags":["story","author_inssein","story_26853339","show_hn"],"author":"inssein","children":[26853341,26854683,26854907,26854924,26854966,26855365,26855755,26855825,26856246,26856632,26857006,26858466,26858619,26858719,26859656],"created_at":"2021-04-18T15:32:46Z","created_at_i":1618759966,"num_comments":41,"objectID":"26853339","points":97,"story_id":26853339,"story_text":"Hi All,
I\u2019ve been working on Aidmin for quite some time now, and finally released a beta version of it.
I\u2019ve focused most of my efforts on security, making sure that Aidmin can\u2019t negatively impact your database. I talk about that a little bit in the Security Overview (https://github.com/aidmin-io/docs/blob/main/security-overvie...).
I would love to know if Aidmin would be helpful at your workplace, and if not, why? Are there any features that are missing which would make it much more useful?
Thank you for checking out the project!","title":"Show HN: Aidmin \u2013 Web-based database GUI","updated_at":"2024-09-20T08:24:03Z","url":"https://aidmin.io"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"leephillips"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"With the rise of AI, web crawlers are suddenly controversial"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://www.theverge.com/24067997/robots-txt-ai-text-file-web-crawlers-spiders"}},"_tags":["story","author_leephillips","story_39420845"],"author":"leephillips","children":[39421110,39421157,39421221,39421243,39421346,39421507,39421587,39421610,39421620,39421919,39422521,39426161],"created_at":"2024-02-18T17:14:21Z","created_at_i":1708276461,"num_comments":84,"objectID":"39420845","points":90,"story_id":39420845,"title":"With the rise of AI, web crawlers are suddenly controversial","updated_at":"2026-02-04T01:03:23Z","url":"https://www.theverge.com/24067997/robots-txt-ai-text-file-web-crawlers-spiders"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jochenmadler"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Hey HN! We\u2019re Vincent and Jochen from Sitefire (https://sitefire.ai). We have been working together for years, with backgrounds in RL/optimization at Stanford and software engineering from Technical University Munich (TUM).
With Sitefire (YC W26), we help marketing teams get recommended by AI Search (ChatGPT, Google AI Overviews, AI Mode, Claude, etc.). Our software monitors prompts, sees which web pages get cited, and uses these insights to help marketing teams take action, e.g. create YouTube videos or write the right blog posts.
This means we have a commercial stake in AI-generated web content. And for now, high-information, AI-generated content works great to get cited and recommended in AI Search.
But after talking to hundreds of marketing teams, it became clear that everyone despises AI-generated content (\u201cAI slop\u201d). And yet, everyone still wants to leverage AI to create content. So we asked ourselves: what characterizes AI slop? Can we train a model to identify it from human-generated web pages?
Researchers from the University of Maryland and Google DeepMind already asked this question for fiction. Their paper StoryScope (Russell et al., 2026) showed that you can tell AI-written stories from human ones by their structure alone, without looking at the words.
We ported their pipeline to commercial web pages. Using the Wayback Machine, we collected 2,250 blog posts from 268 B2B company websites that were written before ChatGPT existed. For each blog post, five AI models (GPT-5.4, Claude Sonnet 4.6, Gemini 3 Flash, DeepSeek V3.2, Kimi K2.5) wrote their own version.
Instead of looking at the words, we looked at how each post is built. We had an AI model answer 214 questions about every post, e.g. how hard it pushes its own product, whether it backs up its claims with sources, or whether it quotes a named expert. Then we trained a classifier on these answers.
On blog posts it had never seen before, our classifier told AI-generated and human posts apart with 98% accuracy, getting only 19 of 1,740 wrong.
Why does it work so well? Because all five AI models write in a similar shape. Mapping every AI model\u2019s values for these features, we see they cluster together, while the human values sit apart and spread out much more. Of the 1% most unique blog posts in our data set, 149 are human, only 4 are AI.
So what characterizes AI slop? It tells you the same thing three times. The title already promises what you'll get ("How to Cut Onboarding Time in Half"), the intro lays out what's coming, and the ending says it all again. 77% of the AI posts end by repeating their main point, compared to only 12% of the human posts. We call it the tidy, self-announcing blog post.
Still, each AI model has its own accent. We trained a second classifier to tell which of the five AI models wrote a post, or whether a human did. It picks the right author 79% of the time, where random guessing (1 in 6) would get 17%. Almost all of its mistakes are mix-ups between the AI models, not between human and AI.
The cool thing about structural features is that you can't simply reword your way out of it. We had each AI model rewrite its own posts until, on average, 73% of their original 13-word sequences were gone, and the AI slop classifier still worked just as well.
We're building this into Sitefire: our agents get a structural understanding of text, so the posts they write go deeper and vary the way human writing does.
There's a lot we haven't tested yet, like the myriad of humanizer tools, human rewriting, restructuring a post, or prompting an AI model to explicitly avoid these habits. And our human posts are mostly from 2020 to 2022, while the AI posts were generated in August 2026. Structure can't really tell when a human post was written, but it's still not a same-year comparison.
We published the study with all the figures on arXiv: https://arxiv.org/abs/2609.15369.\nThe code is on GitHub: https://github.com/pulse-energy-eu/slopshape
We're pretty sure your own blog isn't AI slop, is it? We built a checker that runs one of your posts through the ten features from the paper, so you can see for yourself (the full report asks for a work email): https://sitefire.ai/slop-checker.
Think you can tell AI slop from human writing? We also made a little game to see if you can keep up with our model, which gets all five rounds right: https://sitefire.ai/spot-the-slop."},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: Training a model to identify AI web content from structure alone"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://arxiv.org/abs/2609.15369"}},"_tags":["story","author_jochenmadler","story_49800566","show_hn"],"author":"jochenmadler","children":[49800614,49805909,49805967,49806091,49807264,49808244,49811314,49812400,49815747,49828297,49829518,49831321,49840722,49882991,49942772],"created_at":"2026-09-22T13:00:49Z","created_at_i":1790082049,"num_comments":28,"objectID":"49800566","points":74,"story_id":49800566,"story_text":"Hey HN! We\u2019re Vincent and Jochen from Sitefire (https://sitefire.ai). We have been working together for years, with backgrounds in RL/optimization at Stanford and software engineering from Technical University Munich (TUM).
With Sitefire (YC W26), we help marketing teams get recommended by AI Search (ChatGPT, Google AI Overviews, AI Mode, Claude, etc.). Our software monitors prompts, sees which web pages get cited, and uses these insights to help marketing teams take action, e.g. create YouTube videos or write the right blog posts.
This means we have a commercial stake in AI-generated web content. And for now, high-information, AI-generated content works great to get cited and recommended in AI Search.
But after talking to hundreds of marketing teams, it became clear that everyone despises AI-generated content (\u201cAI slop\u201d). And yet, everyone still wants to leverage AI to create content. So we asked ourselves: what characterizes AI slop? Can we train a model to identify it from human-generated web pages?
Researchers from the University of Maryland and Google DeepMind already asked this question for fiction. Their paper StoryScope (Russell et al., 2026) showed that you can tell AI-written stories from human ones by their structure alone, without looking at the words.
We ported their pipeline to commercial web pages. Using the Wayback Machine, we collected 2,250 blog posts from 268 B2B company websites that were written before ChatGPT existed. For each blog post, five AI models (GPT-5.4, Claude Sonnet 4.6, Gemini 3 Flash, DeepSeek V3.2, Kimi K2.5) wrote their own version.
Instead of looking at the words, we looked at how each post is built. We had an AI model answer 214 questions about every post, e.g. how hard it pushes its own product, whether it backs up its claims with sources, or whether it quotes a named expert. Then we trained a classifier on these answers.
On blog posts it had never seen before, our classifier told AI-generated and human posts apart with 98% accuracy, getting only 19 of 1,740 wrong.
Why does it work so well? Because all five AI models write in a similar shape. Mapping every AI model\u2019s values for these features, we see they cluster together, while the human values sit apart and spread out much more. Of the 1% most unique blog posts in our data set, 149 are human, only 4 are AI.
So what characterizes AI slop? It tells you the same thing three times. The title already promises what you'll get ("How to Cut Onboarding Time in Half"), the intro lays out what's coming, and the ending says it all again. 77% of the AI posts end by repeating their main point, compared to only 12% of the human posts. We call it the tidy, self-announcing blog post.
Still, each AI model has its own accent. We trained a second classifier to tell which of the five AI models wrote a post, or whether a human did. It picks the right author 79% of the time, where random guessing (1 in 6) would get 17%. Almost all of its mistakes are mix-ups between the AI models, not between human and AI.
The cool thing about structural features is that you can't simply reword your way out of it. We had each AI model rewrite its own posts until, on average, 73% of their original 13-word sequences were gone, and the AI slop classifier still worked just as well.
We're building this into Sitefire: our agents get a structural understanding of text, so the posts they write go deeper and vary the way human writing does.
There's a lot we haven't tested yet, like the myriad of humanizer tools, human rewriting, restructuring a post, or prompting an AI model to explicitly avoid these habits. And our human posts are mostly from 2020 to 2022, while the AI posts were generated in August 2026. Structure can't really tell when a human post was written, but it's still not a same-year comparison.
We published the study with all the figures on arXiv: https://arxiv.org/abs/2609.15369.\nThe code is on GitHub: https://github.com/pulse-energy-eu/slopshape
We're pretty sure your own blog isn't AI slop, is it? We built a checker that runs one of your posts through the ten features from the paper, so you can see for yourself (the full report asks for a work email): https://sitefire.ai/slop-checker.
Think you can tell AI slop from human writing? We also made a little game to see if you can keep up with our model, which gets all five rounds right: https://sitefire.ai/spot-the-slop.","title":"Show HN: Training a model to identify AI web content from structure alone","updated_at":"2026-10-06T13:37:47Z","url":"https://arxiv.org/abs/2609.15369"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"d4kmor"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"WebAIM Screen Reader User Survey #9"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://webaim.org/projects/screenreadersurvey9/"}},"_tags":["story","author_d4kmor","story_27075359"],"author":"d4kmor","children":[27076954],"created_at":"2021-05-07T12:58:48Z","created_at_i":1620392328,"num_comments":9,"objectID":"27075359","points":48,"story_id":27075359,"title":"WebAIM Screen Reader User Survey #9","updated_at":"2024-09-20T08:35:02Z","url":"https://webaim.org/projects/screenreadersurvey9/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"phs318u"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Australian influencer Lily Jay's tangled web of AI manipulation"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"https://www.abc.net.au/news/2026-07-05/lily-jay-foundation-posts-ai-generated-misleading-videos/106866422"}},"_tags":["story","author_phs318u","story_48789416"],"author":"phs318u","children":[48789658,48790231,48790732,48793538],"created_at":"2026-07-04T21:50:14Z","created_at_i":1783201814,"num_comments":6,"objectID":"48789416","points":45,"story_id":48789416,"title":"Australian influencer Lily Jay's tangled web of AI manipulation","updated_at":"2026-07-08T11:11:12Z","url":"https://www.abc.net.au/news/2026-07-05/lily-jay-foundation-posts-ai-generated-misleading-videos/106866422"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"navanchauhan"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Man hacked airline Web site to locate his lost luggage"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://boingboing.net/2022/03/31/man-hacked-airline-web-site-to-locate-his-lost-luggage-after-customer-service-was-unhelpful.html"}},"_tags":["story","author_navanchauhan","story_30870134"],"author":"navanchauhan","children":[30870792,30870959,30870966,30871118,30871194,30871547,30872588,30872681,30873530],"created_at":"2022-03-31T17:45:51Z","created_at_i":1648748751,"num_comments":14,"objectID":"30870134","points":42,"story_id":30870134,"title":"Man hacked airline Web site to locate his lost luggage","updated_at":"2024-09-20T10:49:38Z","url":"https://boingboing.net/2022/03/31/man-hacked-airline-web-site-to-locate-his-lost-luggage-after-customer-service-was-unhelpful.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"sqiubncwab"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"WebAI Kinda Stinks"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.capsolver.com/"}},"_tags":["story","author_sqiubncwab","story_47306090"],"author":"sqiubncwab","children":[47306091,47306266],"created_at":"2026-03-09T08:08:50Z","created_at_i":1773043730,"num_comments":1,"objectID":"47306090","points":35,"story_id":47306090,"title":"WebAI Kinda Stinks","updated_at":"2026-03-10T23:51:14Z","url":"https://www.capsolver.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tchalla"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Web-scraping AI bots cause disruption for scientific databases and journals"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.nature.com/articles/d41586-025-01661-4"}},"_tags":["story","author_tchalla","story_44241089"],"author":"tchalla","children":[44241849,44242316,44244821],"created_at":"2025-06-10T20:25:01Z","created_at_i":1749587101,"num_comments":25,"objectID":"44241089","points":31,"story_id":44241089,"title":"Web-scraping AI bots cause disruption for scientific databases and journals","updated_at":"2025-12-01T19:19:40Z","url":"https://www.nature.com/articles/d41586-025-01661-4"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"cdme"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Go ahead and block AI web crawlers"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://coryd.dev/posts/2024/go-ahead-and-block-ai-web-crawlers/"}},"_tags":["story","author_cdme","story_39574220"],"author":"cdme","children":[39574317,39576105,39576123,39576271,39576415,39576483,39576829],"created_at":"2024-03-02T17:47:48Z","created_at_i":1709401668,"num_comments":22,"objectID":"39574220","points":26,"story_id":39574220,"title":"Go ahead and block AI web crawlers","updated_at":"2024-09-20T16:32:38Z","url":"https://coryd.dev/posts/2024/go-ahead-and-block-ai-web-crawlers/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rbanffy"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"IBM RXN for Chemistry: AI web service for predicting chemical reactions"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://rxn.res.ibm.com/"}},"_tags":["story","author_rbanffy","story_17818160"],"author":"rbanffy","children":[17824458,17824479],"created_at":"2018-08-22T12:23:39Z","created_at_i":1534940619,"num_comments":3,"objectID":"17818160","points":24,"story_id":17818160,"title":"IBM RXN for Chemistry: AI web service for predicting chemical reactions","updated_at":"2024-09-20T02:54:24Z","url":"https://rxn.res.ibm.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"bluehatbrit"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Abusive AI Web Crawlers: Get Off My Lawn"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://www.mythic-beasts.com/blog/2025/04/01/abusive-ai-web-crawlers-get-off-my-lawn/"}},"_tags":["story","author_bluehatbrit","story_43556521"],"author":"bluehatbrit","children":[43556821,43556832,43557728,43557850,43563342],"created_at":"2025-04-02T13:37:16Z","created_at_i":1743601036,"num_comments":22,"objectID":"43556521","points":23,"story_id":43556521,"title":"Abusive AI Web Crawlers: Get Off My Lawn","updated_at":"2025-10-06T20:24:15Z","url":"https://www.mythic-beasts.com/blog/2025/04/01/abusive-ai-web-crawlers-get-off-my-lawn/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"akerl_"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Introducing Airpal: web-based query execution tool"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"http://nerds.airbnb.com/airpal/"}},"_tags":["story","author_akerl_","story_9151620"],"author":"akerl_","children":[9153067],"created_at":"2015-03-05T16:16:09Z","created_at_i":1425572169,"num_comments":1,"objectID":"9151620","points":17,"story_id":9151620,"story_text":"","title":"Introducing Airpal: web-based query execution tool","updated_at":"2023-09-07T02:14:12Z","url":"http://nerds.airbnb.com/airpal/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"benzguo"},"story_text":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"Hey HN! I'm open-sourcing a little weekend side project. It's a terminal UI that generates parsers for webpages using AI, and a library that runs these parsers while you browse (it connects to Chrome over CDP). You can parse Twitter, LinkedIn, and HN posts and they get saved to a DuckDB database, so you can query them later. Contributions welcome!
I wrote a bit more about it here: https://0thernet.substack.com/p/memo-2-selectron"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: Open-source AI web parser lib & TUI"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/SubstrateLabs/selectron"}},"_tags":["story","author_benzguo","story_43895048","show_hn"],"author":"benzguo","children":[43895887],"created_at":"2025-05-05T13:44:03Z","created_at_i":1746452643,"num_comments":1,"objectID":"43895048","points":17,"story_id":43895048,"story_text":"Hey HN! I'm open-sourcing a little weekend side project. It's a terminal UI that generates parsers for webpages using AI, and a library that runs these parsers while you browse (it connects to Chrome over CDP). You can parse Twitter, LinkedIn, and HN posts and they get saved to a DuckDB database, so you can query them later. Contributions welcome!
I wrote a bit more about it here: https://0thernet.substack.com/p/memo-2-selectron","title":"Show HN: Open-source AI web parser lib & TUI","updated_at":"2025-05-17T13:21:38Z","url":"https://github.com/SubstrateLabs/selectron"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"vikp"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"A vision for the AI web: the real web 3.0"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://www.vikas.sh/post/vision-for-ai-web"}},"_tags":["story","author_vikp","story_37038716"],"author":"vikp","children":[37039451,37039624,37039671,37039681,37039730,37039763,37040109],"created_at":"2023-08-07T18:40:27Z","created_at_i":1691433627,"num_comments":10,"objectID":"37038716","points":15,"story_id":37038716,"title":"A vision for the AI web: the real web 3.0","updated_at":"2025-08-15T22:14:44Z","url":"https://www.vikas.sh/post/vision-for-ai-web"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ali_chherawalla"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Nine days ago I posted Off Grid here and you showed up - 124 points, 66 comments, bug reports I fixed same-day, and the kind of feedback that makes open source worth it.
You told me what you wanted. Here's what I shipped:\nYour AI can now use tools \u2014 entirely offline.
Web search, calculator, date/time, device info \u2014 with automatic tool loops.
Your 3B parameter model doesn't just generate text anymore. It reasons, calls tools, and synthesizes results.
On your phone. No API key. No server. No cloud function.
So what? It means the gap between "local toy" and "useful assistant" just got dramatically smaller.
You don't need GPT-4 to look something up and give you an answer. A quantized Qwen 3 / SMOLLM3 running on your Snapdragon can do it in no time.
3x faster with configurable KV cache.\nYou can now choose between f16, q8_0, and q4_0 KV cache types. On q4_0, models that were doing 10 tok/s are hitting 30. The app even nudges you after your first generation: "Hey, you could be running faster." One tap.
So what? The #1 complaint about on-device AI is "it's too slow to be useful." That argument just lost a lot of weight. 30tokens/second on a phone is faster than most people read.
Live on both stores. No sideloading. No Xcode.
Off Grid is now on the App Store and Google Play. Install it like any other app. Your parents could use this.
So what? On-device AI just went from "cool weekend project for developers" to "thing normal people can actually try." That matters because privacy shouldn't require a CS degree.
What hasn't changed:\n- MIT licensed. Fully open source. Every line\n- Zero data leaves your device. No analytics. No telemetry. No "anonymous usage data." \n- Text gen (15-30 tok/s), image gen (5-10s on NPU), vision AI, voice transcription, document analysis \n\u2014 all offline\n- Bring any GGUF model. Run Qwen 3, Llama 3.2, Gemma 3, Phi-4, whatever you want.
I'm building this because I believe the phone in your pocket should be the most private computer you own \u2014 not the most surveilled. Every week the models get smaller and faster. The hardware is already there. The software just needs to catch up.
https://github.com/alichherawalla/off-grid-mobile
I'm in the comments. Tell me what to build next."},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: Off Grid: On-device AI-web browsing, tools vision,image,voice\u20133x faster"}},"_tags":["story","author_ali_chherawalla","story_47142003","show_hn"],"author":"ali_chherawalla","children":[47142780,47142863,47144024,47145717,47147901,47160361],"created_at":"2026-02-24T19:57:36Z","created_at_i":1771963056,"num_comments":5,"objectID":"47142003","points":12,"story_id":47142003,"story_text":"Nine days ago I posted Off Grid here and you showed up - 124 points, 66 comments, bug reports I fixed same-day, and the kind of feedback that makes open source worth it.
You told me what you wanted. Here's what I shipped:\nYour AI can now use tools \u2014 entirely offline.
Web search, calculator, date/time, device info \u2014 with automatic tool loops.
Your 3B parameter model doesn't just generate text anymore. It reasons, calls tools, and synthesizes results.
On your phone. No API key. No server. No cloud function.
So what? It means the gap between "local toy" and "useful assistant" just got dramatically smaller.
You don't need GPT-4 to look something up and give you an answer. A quantized Qwen 3 / SMOLLM3 running on your Snapdragon can do it in no time.
3x faster with configurable KV cache.\nYou can now choose between f16, q8_0, and q4_0 KV cache types. On q4_0, models that were doing 10 tok/s are hitting 30. The app even nudges you after your first generation: "Hey, you could be running faster." One tap.
So what? The #1 complaint about on-device AI is "it's too slow to be useful." That argument just lost a lot of weight. 30tokens/second on a phone is faster than most people read.
Live on both stores. No sideloading. No Xcode.
Off Grid is now on the App Store and Google Play. Install it like any other app. Your parents could use this.
So what? On-device AI just went from "cool weekend project for developers" to "thing normal people can actually try." That matters because privacy shouldn't require a CS degree.
What hasn't changed:\n- MIT licensed. Fully open source. Every line\n- Zero data leaves your device. No analytics. No telemetry. No "anonymous usage data." \n- Text gen (15-30 tok/s), image gen (5-10s on NPU), vision AI, voice transcription, document analysis \n\u2014 all offline\n- Bring any GGUF model. Run Qwen 3, Llama 3.2, Gemma 3, Phi-4, whatever you want.
I'm building this because I believe the phone in your pocket should be the most private computer you own \u2014 not the most surveilled. Every week the models get smaller and faster. The hardware is already there. The software just needs to catch up.
https://github.com/alichherawalla/off-grid-mobile
I'm in the comments. Tell me what to build next.","title":"Show HN: Off Grid: On-device AI-web browsing, tools vision,image,voice\u20133x faster","updated_at":"2026-03-25T16:42:23Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"div9"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Multi\u00b7On: Your AI Web Co-Pilot Powered by ChatGPT"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"https://multion.ai"}},"_tags":["story","author_div9","story_34553790"],"author":"div9","children":[34553791,34554615],"created_at":"2023-01-28T00:41:23Z","created_at_i":1674866483,"num_comments":4,"objectID":"34553790","points":12,"story_id":34553790,"title":"Multi\u00b7On: Your AI Web Co-Pilot Powered by ChatGPT","updated_at":"2024-09-20T13:13:18Z","url":"https://multion.ai"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"sweetdreams"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Iran's Web Spying Aided By Western Technology"},"url":{"matchLevel":"none","matchedWords":[],"value":"http://online.wsj.com/article/SB124562668777335653.html"}},"_tags":["story","author_sweetdreams","story_667984"],"author":"sweetdreams","children":[669115],"created_at":"2009-06-22T00:19:28Z","created_at_i":1245629968,"num_comments":1,"objectID":"667984","points":12,"story_id":667984,"title":"Iran's Web Spying Aided By Western Technology","updated_at":"2024-09-19T16:41:49Z","url":"http://online.wsj.com/article/SB124562668777335653.html"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ukdm"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Google's new buggy microblogging Web app aims to educate"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"http://arstechnica.com/web/news/2010/05/googles-new-buggy-microblogging-web-app-aims-to-educate.ars"}},"_tags":["story","author_ukdm","story_1321257"],"author":"ukdm","created_at":"2010-05-05T14:21:05Z","created_at_i":1273069265,"num_comments":0,"objectID":"1321257","points":11,"story_id":1321257,"story_text":"","title":"Google's new buggy microblogging Web app aims to educate","updated_at":"2024-09-19T16:58:28Z","url":"http://arstechnica.com/web/news/2010/05/googles-new-buggy-microblogging-web-app-aims-to-educate.ars"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"gsky"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Cloudflare debuts one-click nuke of web-scraping AI"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"https://www.theregister.com/2024/07/03/cloudflare_ai_blocks/"}},"_tags":["story","author_gsky","story_40870213"],"author":"gsky","children":[40870325,40870359,40870598],"created_at":"2024-07-03T21:35:34Z","created_at_i":1720042534,"num_comments":5,"objectID":"40870213","points":10,"story_id":40870213,"title":"Cloudflare debuts one-click nuke of web-scraping AI","updated_at":"2024-09-20T17:25:21Z","url":"https://www.theregister.com/2024/07/03/cloudflare_ai_blocks/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"colesantiago"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Anubis \u2013 Open-Source Web AI Firewall Utility"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/TecharoHQ/anubis"}},"_tags":["story","author_colesantiago","story_44477785"],"author":"colesantiago","children":[44479552,44483357,44527533],"created_at":"2025-07-06T04:14:27Z","created_at_i":1751775267,"num_comments":3,"objectID":"44477785","points":10,"story_id":44477785,"title":"Anubis \u2013 Open-Source Web AI Firewall Utility","updated_at":"2025-07-16T17:43:16Z","url":"https://github.com/TecharoHQ/anubis"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"amyWang419"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"BrainyAI is a free Chrome browser extension that offers features such as AI chat aggregation, AI search, AI reading, and enhanced AI web browsing. (Free alternative to Sider and Monica)."},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: All-in-one AI web assistant (open-source)"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/luyu0279/BrainyAI"}},"_tags":["story","author_amyWang419","story_40772342","show_hn"],"author":"amyWang419","children":[40772648],"created_at":"2024-06-24T02:49:56Z","created_at_i":1719197396,"num_comments":3,"objectID":"40772342","points":10,"story_id":40772342,"story_text":"BrainyAI is a free Chrome browser extension that offers features such as AI chat aggregation, AI search, AI reading, and enhanced AI web browsing. (Free alternative to Sider and Monica).","title":"Show HN: All-in-one AI web assistant (open-source)","updated_at":"2024-09-20T17:19:07Z","url":"https://github.com/luyu0279/BrainyAI"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"cmonday"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: Fuji-Web \u2013 An Open-Source AI Web Agent with SoTA Performance"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["web"],"value":"https://github.com/normal-computing/fuji-web"}},"_tags":["story","author_cmonday","story_40458478","show_hn"],"author":"cmonday","children":[40458502,40460356,40466976],"created_at":"2024-05-23T18:56:47Z","created_at_i":1716490607,"num_comments":3,"objectID":"40458478","points":10,"story_id":40458478,"title":"Show HN: Fuji-Web \u2013 An Open-Source AI Web Agent with SoTA Performance","updated_at":"2024-09-20T17:05:36Z","url":"https://github.com/normal-computing/fuji-web"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Bluestein"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Google is killing the web with AI Overviews \u2013 I made an extension to block them"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"https://www.tomshardware.com/how-to/block-google-ai-overviews"}},"_tags":["story","author_Bluestein","story_44346440"],"author":"Bluestein","created_at":"2025-06-22T12:41:23Z","created_at_i":1750596083,"num_comments":0,"objectID":"44346440","points":10,"story_id":44346440,"title":"Google is killing the web with AI Overviews \u2013 I made an extension to block them","updated_at":"2025-06-24T07:48:50Z","url":"https://www.tomshardware.com/how-to/block-google-ai-overviews"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Terretta"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Cloudflare's CEO wants to save the web from AI's oligarchs"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["web"],"value":"https://crazystupidtech.com/2025/08/30/cloudflares-ceo-wants-to-save-the-web-from-ais-oligarchs-heres-why-his-plan-isnt-crazy/"}},"_tags":["story","author_Terretta","story_45093275"],"author":"Terretta","children":[45093288,45098108],"created_at":"2025-09-01T15:08:33Z","created_at_i":1756739313,"num_comments":4,"objectID":"45093275","points":9,"story_id":45093275,"title":"Cloudflare's CEO wants to save the web from AI's oligarchs","updated_at":"2026-03-05T22:36:03Z","url":"https://crazystupidtech.com/2025/08/30/cloudflares-ceo-wants-to-save-the-web-from-ais-oligarchs-heres-why-his-plan-isnt-crazy/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"arjunchint"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"We just benchmarked our agent, rtrvr.ai, on the Halluminate (YC S25) Web Bench, and rtrvr.ai achieved a new State-of-the-Art performance with an 81% success rate. For perspective, this surpasses not only all other autonomous agents but also the human-intervention baseline of OpenAI's Operator (76.5%).
It also completes tasks an astonishing 7x faster than the next leading alternative.
This isn't just an incremental improvement; it's a validation of our core architectural philosophy. Our performance stems from two key differentiators:
- Local-First Operation: As a Chrome Extension, rtrvr.ai operates directly within the user's browser. This eliminates the latency, bot detection and access issues that plague cloud browser agents.
- DOM-Based Interaction: Instead of relying on brittle visual parsing (CUA), our agent interacts directly with the page's HTML structure, enabling skipping clicks and resilience to pop-ups and overlays. We also can just use the latest and fastest models such as Gemini Flash for superior performance.
This leads to a critical industry insight: Cloud Browser Agents are not a viable long-term solution for reliable web automation.
Our benchmark analysis shows that over 94% of rtrvr.ai's failures were "agent errors" (fixable AI logic), while only 5% were "infrastructure errors." For cloud agents, this ratio is often inverted. You can't build a reliable agent if you can't even guarantee access to the environment.
Finally it only cost us ~$40 to run this benchmark, whereas we estimate it cost >~$1k in infra costs for each agent for Halluminate.
The future of web automation won't be fought from remote data centers. It will be run symbiotically from your browser. Our results are the first major data point proving this thesis and putting the first nail in the coffin for cloud browser agents.
Full Report: https://www.rtrvr.ai/blog/web-bench-results
Or if you just want to tune into some Agentic-SMR of a web agent doing tasks online tune into the playlist: https://www.youtube.com/watch?v=HWPZI8PjuLY&list=PL5rk1YARPB...
Try out the magic of a working web agent yourself, install at: https://chromewebstore.google.com/detail/rtrvrai-ai-web-agen...
Bring your own API Key from ai.studio and use Google's Gemini Free Tier to use our web agent for free! We literally have a button that will get our agent to open AI Studio create key and configure itself all automatically."},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: rtrvr.ai \u2013 New Free SOTA AI Web Agent Beats Even Operator"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://www.rtrvr.ai/blog/web-bench-results"}},"_tags":["story","author_arjunchint","story_44349797","show_hn"],"author":"arjunchint","children":[44350005],"created_at":"2025-06-22T19:57:01Z","created_at_i":1750622221,"num_comments":2,"objectID":"44349797","points":8,"story_id":44349797,"story_text":"We just benchmarked our agent, rtrvr.ai, on the Halluminate (YC S25) Web Bench, and rtrvr.ai achieved a new State-of-the-Art performance with an 81% success rate. For perspective, this surpasses not only all other autonomous agents but also the human-intervention baseline of OpenAI's Operator (76.5%).
It also completes tasks an astonishing 7x faster than the next leading alternative.
This isn't just an incremental improvement; it's a validation of our core architectural philosophy. Our performance stems from two key differentiators:
- Local-First Operation: As a Chrome Extension, rtrvr.ai operates directly within the user's browser. This eliminates the latency, bot detection and access issues that plague cloud browser agents.
- DOM-Based Interaction: Instead of relying on brittle visual parsing (CUA), our agent interacts directly with the page's HTML structure, enabling skipping clicks and resilience to pop-ups and overlays. We also can just use the latest and fastest models such as Gemini Flash for superior performance.
This leads to a critical industry insight: Cloud Browser Agents are not a viable long-term solution for reliable web automation.
Our benchmark analysis shows that over 94% of rtrvr.ai's failures were "agent errors" (fixable AI logic), while only 5% were "infrastructure errors." For cloud agents, this ratio is often inverted. You can't build a reliable agent if you can't even guarantee access to the environment.
Finally it only cost us ~$40 to run this benchmark, whereas we estimate it cost >~$1k in infra costs for each agent for Halluminate.
The future of web automation won't be fought from remote data centers. It will be run symbiotically from your browser. Our results are the first major data point proving this thesis and putting the first nail in the coffin for cloud browser agents.
Full Report: https://www.rtrvr.ai/blog/web-bench-results
Or if you just want to tune into some Agentic-SMR of a web agent doing tasks online tune into the playlist: https://www.youtube.com/watch?v=HWPZI8PjuLY&list=PL5rk1YARPB...
Try out the magic of a working web agent yourself, install at: https://chromewebstore.google.com/detail/rtrvrai-ai-web-agen...
Bring your own API Key from ai.studio and use Google's Gemini Free Tier to use our web agent for free! We literally have a button that will get our agent to open AI Studio create key and configure itself all automatically.","title":"Show HN: rtrvr.ai \u2013 New Free SOTA AI Web Agent Beats Even Operator","updated_at":"2025-07-14T11:57:38Z","url":"https://www.rtrvr.ai/blog/web-bench-results"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"gscott"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"An 'uncensored' dark web AI is offering step-by-step guides on how to make meth"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://www.dailydot.com/debug/oniongpt-uncesored-dark-web-ai-chatbot/"}},"_tags":["story","author_gscott","story_42612324"],"author":"gscott","children":[42612479,42613746],"created_at":"2025-01-06T16:58:14Z","created_at_i":1736182694,"num_comments":2,"objectID":"42612324","points":8,"story_id":42612324,"title":"An 'uncensored' dark web AI is offering step-by-step guides on how to make meth","updated_at":"2025-01-08T17:54:31Z","url":"https://www.dailydot.com/debug/oniongpt-uncesored-dark-web-ai-chatbot/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tambourine_man"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Copilot for the Web: AI-Powered Bing and Microsoft Edge"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/watch?v=rOeRWRJ16yY"}},"_tags":["story","author_tambourine_man","story_34715411"],"author":"tambourine_man","children":[34715490,34716619],"created_at":"2023-02-08T21:24:48Z","created_at_i":1675891488,"num_comments":2,"objectID":"34715411","points":8,"story_id":34715411,"title":"Copilot for the Web: AI-Powered Bing and Microsoft Edge","updated_at":"2024-09-20T13:11:30Z","url":"https://www.youtube.com/watch?v=rOeRWRJ16yY"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Bender"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Cloudflare Puts a Default Block on AI Web Scraping"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://www.securityweek.com/cloudflare-puts-a-default-block-on-ai-web-scraping/"}},"_tags":["story","author_Bender","story_44490378"],"author":"Bender","children":[44490484],"created_at":"2025-07-07T13:51:09Z","created_at_i":1751896269,"num_comments":0,"objectID":"44490378","points":8,"story_id":44490378,"title":"Cloudflare Puts a Default Block on AI Web Scraping","updated_at":"2025-07-08T15:25:16Z","url":"https://www.securityweek.com/cloudflare-puts-a-default-block-on-ai-web-scraping/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rntn"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"AI killing the old web: Generative AI models are changing the economy of the web"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://www.theverge.com/2023/6/26/23773914/ai-large-language-models-data-scraping-generation-remaking-web"}},"_tags":["story","author_rntn","story_36483570"],"author":"rntn","created_at":"2023-06-26T18:19:08Z","created_at_i":1687803548,"num_comments":0,"objectID":"36483570","points":8,"story_id":36483570,"title":"AI killing the old web: Generative AI models are changing the economy of the web","updated_at":"2024-09-20T14:30:07Z","url":"https://www.theverge.com/2023/6/26/23773914/ai-large-language-models-data-scraping-generation-remaking-web"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"arjunchint"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Hey HN,
I'm excited to share rtrvr.ai, a Chrome extension that brings the power of AI agents to your everyday web browsing. It's designed to automate complex web tasks, extract structured data from any website, and integrate with your favorite tools as you browse using AI Function Calling [ie: \u201cSend this page summary as Slack message\u201d].
The core idea is to let anyone, even non-developers, leverage the power of web automation and data extraction using natural language. Imagine being able to:
Automate lead generation: Extract hundreds of LinkedIn profiles to Google Sheets, complete with AI-generated, personalized intro emails.
Process PDFs in bulk: Pull data like revenue, expenses, and totals from hundreds of local or online PDFs directly into your spreadsheets.
Navigate and extract from paginated lists: Tell rtrvr.ai, "For each YC Partner, go to their LinkedIn profile and retrieve their name, headline, job, and college," and it'll do it.
Automate workflows across multiple tabs: For example, fill out job applications on Careers tab using information from the LinkedIn tab.
Function Calling: Integrate with APIs like Snowflake and Slack directly from your prompts using simple @ notation or letting the AI infer what tool to use with natural language.
GraphBot: Generate charts and visualizations from website data with natural language commands.
Recordings: Ground the agent with site interactions recordings to ensure accurate and repeatable task execution.
Sheet Context: Use Google Sheets data as context for your web tasks.
Scheduling: Run automations on a schedule in the background.
Sheets Workflow: You can feed a Google Sheet with a column of URLs (like LinkedIn profiles), and it will open each url as a tab, extract data, and even generate content (like intro emails) back into the sheet. It can handle multi-step workflows with prior output dependencies, effectively representing a DAG.
I see rtrvr.ai as a step towards a more intelligent and interactive web. I believe this tool can be a game-changer for marketers, sales professionals, SMBs, and anyone who needs to extract information from the web efficiently. Would love to hear your feedback, suggestions, and any use cases you can think of!
Website: rtrvr.ai\nChrome Store Listing: https://chromewebstore.google.com/detail/rtrvrai/jldogdgepmc...\nYoutube Intro: https://www.youtube.com/watch?v=wajCM6208cc"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: rtrvr.ai \u2013 AI Web Agent for Automating Workflows and Data Extraction"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://chromewebstore.google.com/detail/rtrvrai/jldogdgepmcedfdhgnmclgemehfhpomg"}},"_tags":["story","author_arjunchint","story_42496918","show_hn"],"author":"arjunchint","children":[42498482,42505275,42528238],"created_at":"2024-12-23T19:22:24Z","created_at_i":1734981744,"num_comments":4,"objectID":"42496918","points":7,"story_id":42496918,"story_text":"Hey HN,
I'm excited to share rtrvr.ai, a Chrome extension that brings the power of AI agents to your everyday web browsing. It's designed to automate complex web tasks, extract structured data from any website, and integrate with your favorite tools as you browse using AI Function Calling [ie: \u201cSend this page summary as Slack message\u201d].
The core idea is to let anyone, even non-developers, leverage the power of web automation and data extraction using natural language. Imagine being able to:
Automate lead generation: Extract hundreds of LinkedIn profiles to Google Sheets, complete with AI-generated, personalized intro emails.
Process PDFs in bulk: Pull data like revenue, expenses, and totals from hundreds of local or online PDFs directly into your spreadsheets.
Navigate and extract from paginated lists: Tell rtrvr.ai, "For each YC Partner, go to their LinkedIn profile and retrieve their name, headline, job, and college," and it'll do it.
Automate workflows across multiple tabs: For example, fill out job applications on Careers tab using information from the LinkedIn tab.
Function Calling: Integrate with APIs like Snowflake and Slack directly from your prompts using simple @ notation or letting the AI infer what tool to use with natural language.
GraphBot: Generate charts and visualizations from website data with natural language commands.
Recordings: Ground the agent with site interactions recordings to ensure accurate and repeatable task execution.
Sheet Context: Use Google Sheets data as context for your web tasks.
Scheduling: Run automations on a schedule in the background.
Sheets Workflow: You can feed a Google Sheet with a column of URLs (like LinkedIn profiles), and it will open each url as a tab, extract data, and even generate content (like intro emails) back into the sheet. It can handle multi-step workflows with prior output dependencies, effectively representing a DAG.
I see rtrvr.ai as a step towards a more intelligent and interactive web. I believe this tool can be a game-changer for marketers, sales professionals, SMBs, and anyone who needs to extract information from the web efficiently. Would love to hear your feedback, suggestions, and any use cases you can think of!
Website: rtrvr.ai\nChrome Store Listing: https://chromewebstore.google.com/detail/rtrvrai/jldogdgepmc...\nYoutube Intro: https://www.youtube.com/watch?v=wajCM6208cc","title":"Show HN: rtrvr.ai \u2013 AI Web Agent for Automating Workflows and Data Extraction","updated_at":"2025-11-17T21:11:03Z","url":"https://chromewebstore.google.com/detail/rtrvrai/jldogdgepmcedfdhgnmclgemehfhpomg"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"gametorch"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Increased Traffic from Web-Scraping AI Bots Is Hard to Monetize"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://m.slashdot.org/story/443217"}},"_tags":["story","author_gametorch","story_44278918"],"author":"gametorch","children":[44278983,44279012],"created_at":"2025-06-14T21:20:01Z","created_at_i":1749936001,"num_comments":1,"objectID":"44278918","points":7,"story_id":44278918,"title":"Increased Traffic from Web-Scraping AI Bots Is Hard to Monetize","updated_at":"2025-08-14T22:15:22Z","url":"https://m.slashdot.org/story/443217"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"obetomuniz"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"web-ai-sdk: experimenting with browser-native AI APIs and WebMCP"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://web-ai-sdk.dev/"}},"_tags":["story","author_obetomuniz","story_48229072"],"author":"obetomuniz","children":[48229073],"created_at":"2026-05-21T21:27:29Z","created_at_i":1779398849,"num_comments":0,"objectID":"48229072","points":7,"story_id":48229072,"title":"web-ai-sdk: experimenting with browser-native AI APIs and WebMCP","updated_at":"2026-05-22T16:11:27Z","url":"https://web-ai-sdk.dev/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"mcstempel"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"How to block AI web crawlers"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://stytch.com/blog/how-to-block-ai-web-crawlers/"}},"_tags":["story","author_mcstempel","story_44061587"],"author":"mcstempel","children":[44061656],"created_at":"2025-05-22T12:55:35Z","created_at_i":1747918535,"num_comments":0,"objectID":"44061587","points":7,"story_id":44061587,"title":"How to block AI web crawlers","updated_at":"2025-05-22T16:08:16Z","url":"https://stytch.com/blog/how-to-block-ai-web-crawlers/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Ahmed_jemaa"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"I'm seeing a lot of YC-backed companies in the last 2-3 batches doing some sort of ai-augmented web scraping (from Talent sourcing, to CRM enrichment, to M&A Due Diligence, etc..).
I'm curious what's the underlying tech behind them. This can't just be regular web scraping or browser search ?
Example of companies: \nhttps://octolane.com\nhttps://openmart.ai\nhttps://watto.ai\nhttps://topo.io"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"AI Web Scraping?"}},"_tags":["story","author_Ahmed_jemaa","story_39894495","ask_hn"],"author":"Ahmed_jemaa","created_at":"2024-04-01T14:27:30Z","created_at_i":1711981650,"num_comments":0,"objectID":"39894495","points":7,"story_id":39894495,"story_text":"I'm seeing a lot of YC-backed companies in the last 2-3 batches doing some sort of ai-augmented web scraping (from Talent sourcing, to CRM enrichment, to M&A Due Diligence, etc..).
I'm curious what's the underlying tech behind them. This can't just be regular web scraping or browser search ?
Example of companies: \nhttps://octolane.com\nhttps://openmart.ai\nhttps://watto.ai\nhttps://topo.io","title":"AI Web Scraping?","updated_at":"2024-09-20T16:46:44Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"arjunchint"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Hey HN, I'm Arjun, one of the founders of rtrvr.ai.
For the past year, we've been building and iterating on an in-browser AI agent. But a few months ago, we hit a wall in our approach and decided to tear everything down and rebuild from first principles. The result is what we're re-launching today.
The Problem/Insight:\nWe kept coming back to a simple, nagging question: web agents by themselves are a nice to have but we kept getting requests on possible ways to integrate or call our agent from another agent. Users also kept asking to reuse their own existing chatbot subscriptions to power the agent.
The "Aha!" Moment with MCP:\nThis led us down the rabbit hole of interoperability and we landed on MCP. Our core insight was this: what if we exposed the Chrome Extension itself as a Remote MCP server?
This simple idea completely changed our architecture and value proposition. It turns the user's browser into a controllable, sandboxed endpoint that any other AI can talk to with a simple copy/paste of a MCP url.
What this new architecture unlocks:\nRemote Control Your Browser: You can stay in your flow in Claude and tell it to perform tasks on your behalf in background tabs. No more context switching. Tell Claude to file that Jira ticket or scrape a site without ever leaving the chat.
Reuse Your Own Subscription (The "BYO-Sub" model): Since Claude is driving the action, you're using your existing subscription to power the agent for basic tasks. We see this as the future \u2013 our agent provides the secure execution rails, and you bring the brain you already pay for. Users can then upgrade to our more advanced, multi-step agent for complex workflows when needed.
Our Bigger Bet - The Agentic Web:\nWe believe the future isn't a single "god agent" but a network of specialized agents that collaborate. By exposing the browser as a common execution layer, we're hoping to provide the "rails" for this future. We're already in talks with other agent platforms to build on this, enabling a more open and interconnected ecosystem.
We just launched on Product Hunt today and would genuinely love to hear the HN community's feedback. What are the security implications we haven't thought of? What are the unlocked use cases you see by giving an agent/app browser context? What's the next logical step for an open "Agentic Web"? All criticism and ideas are welcome.
Thanks for reading!
Links:\nProduct Hunt Launch: https://www.producthunt.com/products/rtrvr-ai?launch=rtrvr-a...
Chrome Extension: https://chromewebstore.google.com/detail/rtrvrai-ai-web-agen..."},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: Rtrvr.ai \u2013 Remote Control Your Browser with an AI Web Agent and MCP"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"https://www.rtrvr.ai"}},"_tags":["story","author_arjunchint","story_45898043","show_hn"],"author":"arjunchint","children":[45898115,45899922,45908564],"created_at":"2025-11-12T09:16:59Z","created_at_i":1762939019,"num_comments":3,"objectID":"45898043","points":6,"story_id":45898043,"story_text":"Hey HN, I'm Arjun, one of the founders of rtrvr.ai.
For the past year, we've been building and iterating on an in-browser AI agent. But a few months ago, we hit a wall in our approach and decided to tear everything down and rebuild from first principles. The result is what we're re-launching today.
The Problem/Insight:\nWe kept coming back to a simple, nagging question: web agents by themselves are a nice to have but we kept getting requests on possible ways to integrate or call our agent from another agent. Users also kept asking to reuse their own existing chatbot subscriptions to power the agent.
The "Aha!" Moment with MCP:\nThis led us down the rabbit hole of interoperability and we landed on MCP. Our core insight was this: what if we exposed the Chrome Extension itself as a Remote MCP server?
This simple idea completely changed our architecture and value proposition. It turns the user's browser into a controllable, sandboxed endpoint that any other AI can talk to with a simple copy/paste of a MCP url.
What this new architecture unlocks:\nRemote Control Your Browser: You can stay in your flow in Claude and tell it to perform tasks on your behalf in background tabs. No more context switching. Tell Claude to file that Jira ticket or scrape a site without ever leaving the chat.
Reuse Your Own Subscription (The "BYO-Sub" model): Since Claude is driving the action, you're using your existing subscription to power the agent for basic tasks. We see this as the future \u2013 our agent provides the secure execution rails, and you bring the brain you already pay for. Users can then upgrade to our more advanced, multi-step agent for complex workflows when needed.
Our Bigger Bet - The Agentic Web:\nWe believe the future isn't a single "god agent" but a network of specialized agents that collaborate. By exposing the browser as a common execution layer, we're hoping to provide the "rails" for this future. We're already in talks with other agent platforms to build on this, enabling a more open and interconnected ecosystem.
We just launched on Product Hunt today and would genuinely love to hear the HN community's feedback. What are the security implications we haven't thought of? What are the unlocked use cases you see by giving an agent/app browser context? What's the next logical step for an open "Agentic Web"? All criticism and ideas are welcome.
Thanks for reading!
Links:\nProduct Hunt Launch: https://www.producthunt.com/products/rtrvr-ai?launch=rtrvr-a...
Chrome Extension: https://chromewebstore.google.com/detail/rtrvrai-ai-web-agen...","title":"Show HN: Rtrvr.ai \u2013 Remote Control Your Browser with an AI Web Agent and MCP","updated_at":"2026-03-05T23:02:23Z","url":"https://www.rtrvr.ai"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jiayuanzhang"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Search the Web Before AI"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://pregptsearch.com"}},"_tags":["story","author_jiayuanzhang","story_45136054"],"author":"jiayuanzhang","children":[45136248,45136256],"created_at":"2025-09-05T07:51:24Z","created_at_i":1757058684,"num_comments":2,"objectID":"45136054","points":6,"story_id":45136054,"title":"Search the Web Before AI","updated_at":"2026-03-05T22:39:09Z","url":"https://pregptsearch.com"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"arjunchint"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Most of us have a list of URLs we need data from (government listings, local business info, pdf directories). Usually, that means hiring a freelancer or paying for an expensive, rigid SaaS.
We built an AI Web Agent platform, rtrvr.ai to make "Vibe Scraping" a thing.
How it works:
1. Upload a Google Sheet with your URLs.
2. Prompt: "Find the email, phone number, and their top 3 services."
3. Watch the AI agents open 50+ browsers at once and fill your sheet in real-time.
It\u2019s powered by a multi-agent system that can take actions, upload files, and crawl through paginations.
Web Agent technology built from the ground up:
End to End Agent: we built a resilient agentic harness with 20+ specialized sub-agents that transforms a single prompt into a complete end-to-end workflow. Turn any prompt into an end to end workflow, and on any site changes the agent adapts.
Dom Intelligence: we perfected a DOM-only web agent approach that represents any webpage as semantic trees guaranteeing zero hallucinations and leveraging the underlying semantic reasoning capabilities of LLMs.
Native Chrome APIs: we built a Chrome Extension to control cloud browsers that runs in the same process as the browser to avoid the bot detection and failure rates of CDP. We further solved the hard problems of interacting with the Shadow DOM and other DOM edge cases.
Cost: We engineered the cost down to $10/mo but you can bring your own Gemini key and proxies to use for nearly FREE. Compare that to the $200+/mo some other lead gen tools like Clay charge.
Use the free browser extension for login walled sites like LinkedIn locally, or the cloud platform for scale on the public web.
We are thinking it can be a great upstream tool to your CRM to generate lists and enrich data.
Curious to hear if this would make your lead generation, scraping, or automation easier or is it missing the mark?"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: Vibe scrape with AI Web Agents, prompt => get data [video]"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/watch?v=ggLDvZKuBlU"}},"_tags":["story","author_arjunchint","story_46610706","show_hn"],"author":"arjunchint","children":[46610783],"created_at":"2026-01-14T00:30:54Z","created_at_i":1768350654,"num_comments":1,"objectID":"46610706","points":6,"story_id":46610706,"story_text":"Most of us have a list of URLs we need data from (government listings, local business info, pdf directories). Usually, that means hiring a freelancer or paying for an expensive, rigid SaaS.
We built an AI Web Agent platform, rtrvr.ai to make "Vibe Scraping" a thing.
How it works:
1. Upload a Google Sheet with your URLs.
2. Prompt: "Find the email, phone number, and their top 3 services."
3. Watch the AI agents open 50+ browsers at once and fill your sheet in real-time.
It\u2019s powered by a multi-agent system that can take actions, upload files, and crawl through paginations.
Web Agent technology built from the ground up:
End to End Agent: we built a resilient agentic harness with 20+ specialized sub-agents that transforms a single prompt into a complete end-to-end workflow. Turn any prompt into an end to end workflow, and on any site changes the agent adapts.
Dom Intelligence: we perfected a DOM-only web agent approach that represents any webpage as semantic trees guaranteeing zero hallucinations and leveraging the underlying semantic reasoning capabilities of LLMs.
Native Chrome APIs: we built a Chrome Extension to control cloud browsers that runs in the same process as the browser to avoid the bot detection and failure rates of CDP. We further solved the hard problems of interacting with the Shadow DOM and other DOM edge cases.
Cost: We engineered the cost down to $10/mo but you can bring your own Gemini key and proxies to use for nearly FREE. Compare that to the $200+/mo some other lead gen tools like Clay charge.
Use the free browser extension for login walled sites like LinkedIn locally, or the cloud platform for scale on the public web.
We are thinking it can be a great upstream tool to your CRM to generate lists and enrich data.
Curious to hear if this would make your lead generation, scraping, or automation easier or is it missing the mark?","title":"Show HN: Vibe scrape with AI Web Agents, prompt => get data [video]","updated_at":"2026-03-05T23:23:16Z","url":"https://www.youtube.com/watch?v=ggLDvZKuBlU"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"zhihuang"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: AI Web Co-Pilot that understands natural language and executes tasks"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"https://chrome.google.com/webstore/detail/imean-generative-ai-meets/kjbmffebjpefbpfokhammhcnknnebhab"}},"_tags":["story","author_zhihuang","story_36888124","show_hn"],"author":"zhihuang","children":[36888125],"created_at":"2023-07-27T02:24:42Z","created_at_i":1690424682,"num_comments":1,"objectID":"36888124","points":6,"story_id":36888124,"title":"Show HN: AI Web Co-Pilot that understands natural language and executes tasks","updated_at":"2024-09-20T14:42:37Z","url":"https://chrome.google.com/webstore/detail/imean-generative-ai-meets/kjbmffebjpefbpfokhammhcnknnebhab"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"conorwade"},"story_text":{"matchLevel":"none","matchedWords":[],"value":""},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Espresso Framework - Scalable Ruby Web Framework aimed at Speed and Simplicity"},"url":{"matchLevel":"none","matchedWords":[],"value":"http://espresso.github.com/"}},"_tags":["story","author_conorwade","story_5251405"],"author":"conorwade","created_at":"2013-02-20T16:39:08Z","created_at_i":1361378348,"num_comments":0,"objectID":"5251405","points":6,"story_id":5251405,"story_text":"","title":"Espresso Framework - Scalable Ruby Web Framework aimed at Speed and Simplicity","updated_at":"2024-09-19T19:17:36Z","url":"http://espresso.github.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"kordlessagain"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Litert.js, Google's High Performance Web AI Inference"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://developers.googleblog.com/litertjs-googles-high-performance-web-ai-inference/"}},"_tags":["story","author_kordlessagain","story_48881217"],"author":"kordlessagain","created_at":"2026-07-12T13:47:50Z","created_at_i":1783864070,"num_comments":0,"objectID":"48881217","points":6,"story_id":48881217,"title":"Litert.js, Google's High Performance Web AI Inference","updated_at":"2026-07-13T18:00:59Z","url":"https://developers.googleblog.com/litertjs-googles-high-performance-web-ai-inference/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"oopismcgoopis"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"For the past few months, I've poured all my energy into building something I'm incredibly passionate about: SamwiseOS, a completely in-browser, AI-first operating system.
What is it?
SamwiseOS is a persistent, single-page web app that looks and feels like a real operating system, complete with a terminal, a robust virtual filesystem (that saves to your browser's IndexedDB), user/group management, and even graphical applications.
The twist? The entire core logic\u2014filesystem, command execution, user management, everything\u2014runs on a Python kernel powered by Pyodide (WebAssembly). The JS frontend acts as a "Stage Manager," handling the UI, sound, and other browser APIs, while the Python kernel is the single source of truth. They talk to each other through a simple effect contract, which is a fancy way of saying they're best friends who communicate really, really well.
Why did I build this?
I wanted to explore what a truly AI-first OS would feel like. Instead of just a command line, you can interact with SamwiseOS through conversation. The gemini command can use system tools to answer questions about your files, forge can generate code for you, and storyboard can analyze a whole directory of code and tell you what it does. It\u2019s like having a brilliant, tireless intern who lives in your browser. It\u2019s the civil servant of operating systems \u2013 for the people, by the people (and AI).
Features I'm especially proud of:
Hybrid Kernel: A robust, sandboxed Python kernel running in WASM, with a nimble JavaScript frontend. It's the best of both worlds!
AI-Powered Shell: Use commands like gemini, chidi, and forge to interact with the OS using natural language.
100+ POSIX-like Commands: We've got everything from ls, grep, and awk to sudo, chmod, and useradd. It's a real, functional environment.
GUI Apps: It's not just a terminal! Use edit for a text/code editor, paint for an ASCII art editor, top for a process viewer, chidi to analyze documents, and even adventure to play text-based games.
Persistence: Your session, files, users, and command history are all saved in IndexedDB, so you can pick up right where you left off.
Multi-User & Permissions: A full-fledged user and group system, including a virtual /etc/sudoers file and sudo capabilities.
The project is entirely self-contained and runs offline.
I've had an absolute blast building this, and I'm bursting with ideas for the future (check out the roadmap in the README!). I would be honored if you'd take a look, poke around the filesystem, and let me know what you think."},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Show HN: SamwiseOS \u2013 A web-based, AI-first OS with a Python kernel in Pyodide"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://samwiseos.neocities.org"}},"_tags":["story","author_oopismcgoopis","story_44937273","show_hn"],"author":"oopismcgoopis","created_at":"2025-08-18T03:57:44Z","created_at_i":1755489464,"num_comments":0,"objectID":"44937273","points":6,"story_id":44937273,"story_text":"For the past few months, I've poured all my energy into building something I'm incredibly passionate about: SamwiseOS, a completely in-browser, AI-first operating system.
What is it?
SamwiseOS is a persistent, single-page web app that looks and feels like a real operating system, complete with a terminal, a robust virtual filesystem (that saves to your browser's IndexedDB), user/group management, and even graphical applications.
The twist? The entire core logic\u2014filesystem, command execution, user management, everything\u2014runs on a Python kernel powered by Pyodide (WebAssembly). The JS frontend acts as a "Stage Manager," handling the UI, sound, and other browser APIs, while the Python kernel is the single source of truth. They talk to each other through a simple effect contract, which is a fancy way of saying they're best friends who communicate really, really well.
Why did I build this?
I wanted to explore what a truly AI-first OS would feel like. Instead of just a command line, you can interact with SamwiseOS through conversation. The gemini command can use system tools to answer questions about your files, forge can generate code for you, and storyboard can analyze a whole directory of code and tell you what it does. It\u2019s like having a brilliant, tireless intern who lives in your browser. It\u2019s the civil servant of operating systems \u2013 for the people, by the people (and AI).
Features I'm especially proud of:
Hybrid Kernel: A robust, sandboxed Python kernel running in WASM, with a nimble JavaScript frontend. It's the best of both worlds!
AI-Powered Shell: Use commands like gemini, chidi, and forge to interact with the OS using natural language.
100+ POSIX-like Commands: We've got everything from ls, grep, and awk to sudo, chmod, and useradd. It's a real, functional environment.
GUI Apps: It's not just a terminal! Use edit for a text/code editor, paint for an ASCII art editor, top for a process viewer, chidi to analyze documents, and even adventure to play text-based games.
Persistence: Your session, files, users, and command history are all saved in IndexedDB, so you can pick up right where you left off.
Multi-User & Permissions: A full-fledged user and group system, including a virtual /etc/sudoers file and sudo capabilities.
The project is entirely self-contained and runs offline.
I've had an absolute blast building this, and I'm bursting with ideas for the future (check out the roadmap in the README!). I would be honored if you'd take a look, poke around the filesystem, and let me know what you think.","title":"Show HN: SamwiseOS \u2013 A web-based, AI-first OS with a Python kernel in Pyodide","updated_at":"2026-03-05T22:30:26Z","url":"https://samwiseos.neocities.org"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"blackrabbit"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Google's new version of AI Web Search"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["web"],"value":"https://blog.google/products/search/web-guide-labs/"}},"_tags":["story","author_blackrabbit","story_44688791"],"author":"blackrabbit","created_at":"2025-07-25T21:37:51Z","created_at_i":1753479471,"num_comments":0,"objectID":"44688791","points":6,"story_id":44688791,"title":"Google's new version of AI Web Search","updated_at":"2025-08-07T11:08:41Z","url":"https://blog.google/products/search/web-guide-labs/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"earcar"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"Why Ruby may be the best language to write your next AI web application"},"url":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["ai"],"value":"https://losangelesaiapps.com/building-ai-applications-in-ruby/"}},"_tags":["story","author_earcar","story_44003286"],"author":"earcar","created_at":"2025-05-16T09:12:46Z","created_at_i":1747386766,"num_comments":0,"objectID":"44003286","points":6,"story_id":44003286,"title":"Why Ruby may be the best language to write your next AI web application","updated_at":"2025-05-29T16:14:11Z","url":"https://losangelesaiapps.com/building-ai-applications-in-ruby/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tysone"},"title":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"With the rise of AI, web crawlers are suddenly controversial"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["web","ai"],"value":"https://www.theverge.com/24067997/robots-txt-ai-text-file-web-crawlers-spiders"}},"_tags":["story","author_tysone","story_39396344"],"author":"tysone","created_at":"2024-02-16T12:58:36Z","created_at_i":1708088316,"num_comments":0,"objectID":"39396344","points":6,"story_id":39396344,"title":"With the rise of AI, web crawlers are suddenly controversial","updated_at":"2024-09-20T16:31:10Z","url":"https://www.theverge.com/24067997/robots-txt-ai-text-file-web-crawlers-spiders"}],"hitsPerPage":50,"nbHits":6701,"nbPages":20,"page":0,"params":"query=Web+AI&tags=story&hitsPerPage=50&advancedSyntax=true&analyticsTags=backend","processingTimeMS":22,"processingTimingsMS":{"_request":{"roundTrip":19},"afterFetch":{"format":{"highlighting":1,"total":2},"merge":{"mergeLoop":{"prepareNextHit":3,"total":3},"total":3},"total":3},"fetch":{"query":2,"scanning":15,"total":18},"total":22},"query":"Web AI","serverTimeMS":25}