I've used agents extensively in my workflows for the better part of the last year - the biggest pain point was always the browser. Every tool out there assumes a human operator with automation bolted on. I wanted to flip that - make the agent the primary driver and give the human a supervisory role.
Enter: Vessel Browser - an Electron-based browser with 40+ MCP-native tools, persistent sessions that survive restarts, semantic page context (agents get structured meaning, not raw HTML), and a supervisor sidepanel where you can watch and control exactly what the agent is doing.
It works as an MCP server with any compatible harness, or use the built-in assistant with integrated chat and BYOK to 8+ providers including custom OAI compatible endpoints.
Install with: npm i @quanta-intellect/vessel-browser"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Vessel Browser \u2013 An open-source browser built for AI agents, not humans"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://quantaintellect.com/"}},"_tags":["story","author_unmodeledtyler","story_47470156","show_hn"],"author":"unmodeledtyler","children":[47470937,47471295,47485388,47560989,47611970],"created_at":"2026-03-21T19:02:57Z","created_at_i":1774119777,"num_comments":5,"objectID":"47470156","points":6,"story_id":47470156,"story_text":"I'm Tyler - the solo operator of Quanta Intellect based in Portland, Oregon. I recently participated in Nous Research's Hermes Agent Hackathon, which is where this project was born.
I've used agents extensively in my workflows for the better part of the last year - the biggest pain point was always the browser. Every tool out there assumes a human operator with automation bolted on. I wanted to flip that - make the agent the primary driver and give the human a supervisory role.
Enter: Vessel Browser - an Electron-based browser with 40+ MCP-native tools, persistent sessions that survive restarts, semantic page context (agents get structured meaning, not raw HTML), and a supervisor sidepanel where you can watch and control exactly what the agent is doing.
It works as an MCP server with any compatible harness, or use the built-in assistant with integrated chat and BYOK to 8+ providers including custom OAI compatible endpoints.
Install with: npm i @quanta-intellect/vessel-browser","title":"Show HN: Vessel Browser \u2013 An open-source browser built for AI agents, not humans","updated_at":"2026-04-02T12:33:32Z","url":"https://quantaintellect.com/"},{"_highlightResult":{"author":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["research"],"value":"lambda-research"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"Companion is a free, open-source web app, featuring a Python REPL environment with an AI Tutor designed to support one\u2019s learning and problem-solving in programming. I am leveraging the Hermes 3 405B model from Nous Research, hosted on Lambda\u2019s Inference API.
It's community-driven, 100% free, and open to all. I\u2019d love your feedback and suggestions. Here's a short video where I demo the tool: https://www.youtube.com/watch?v=4Plt_sh_cIg&ab_channel=Rahul"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Open-Source Python REPL with AI Tutor for Learning and Problem-Solving"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.companionai.dev/"}},"_tags":["story","author_lambda-research","story_41995903","show_hn"],"author":"lambda-research","children":[41999285],"created_at":"2024-10-30T15:16:27Z","created_at_i":1730301387,"num_comments":2,"objectID":"41995903","points":5,"story_id":41995903,"story_text":"Companion is a free, open-source web app, featuring a Python REPL environment with an AI Tutor designed to support one\u2019s learning and problem-solving in programming. I am leveraging the Hermes 3 405B model from Nous Research, hosted on Lambda\u2019s Inference API.
It's community-driven, 100% free, and open to all. I\u2019d love your feedback and suggestions. Here's a short video where I demo the tool: https://www.youtube.com/watch?v=4Plt_sh_cIg&ab_channel=Rahul","title":"Show HN: Open-Source Python REPL with AI Tutor for Learning and Problem-Solving","updated_at":"2024-11-04T10:28:38Z","url":"https://www.companionai.dev/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"ethanjamescolez"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"This is an independent Hermes Agent install guide I put together for the part that usually gets skipped after "run this command."
What I wanted was one place that shows the environment choice first, then the official installer path, then the first commands people usually need right after install. The practical routes here are macOS, Linux, WSL2, and Termux.
The flow is intentionally narrow: choose the host, run the official installer command, reload the shell, then continue into commands like hermes, hermes model, hermes setup, hermes gateway, hermes doctor, and hermes update while keeping the official docs and GitHub close for verification.
AI Hermes Agent is independent, is not the official Hermes Agent website, is not owned or endorsed by Nous Research, does not host or run Hermes Agent, and does not provide an online chat demo, account system, subscription, credits, API access, or managed runtime.
If you try it, I would especially like feedback on whether the environment choice and the post-install steps are clear enough, and where the guide still leaves too much implied."},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: AI Hermes Agent \u2013 An independent Hermes Agent install guide"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://ai-hermes-agent.com/install"}},"_tags":["story","author_ethanjamescolez","story_47865412","show_hn"],"author":"ethanjamescolez","children":[47865659],"created_at":"2026-04-22T15:52:27Z","created_at_i":1776873147,"num_comments":1,"objectID":"47865412","points":3,"story_id":47865412,"story_text":"This is an independent Hermes Agent install guide I put together for the part that usually gets skipped after "run this command."
What I wanted was one place that shows the environment choice first, then the official installer path, then the first commands people usually need right after install. The practical routes here are macOS, Linux, WSL2, and Termux.
The flow is intentionally narrow: choose the host, run the official installer command, reload the shell, then continue into commands like hermes, hermes model, hermes setup, hermes gateway, hermes doctor, and hermes update while keeping the official docs and GitHub close for verification.
AI Hermes Agent is independent, is not the official Hermes Agent website, is not owned or endorsed by Nous Research, does not host or run Hermes Agent, and does not provide an online chat demo, account system, subscription, credits, API access, or managed runtime.
If you try it, I would especially like feedback on whether the environment choice and the post-install steps are clear enough, and where the guide still leaves too much implied.","title":"Show HN: AI Hermes Agent \u2013 An independent Hermes Agent install guide","updated_at":"2026-05-02T05:21:41Z","url":"https://ai-hermes-agent.com/install"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"GravyPouch"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"I felt like ChatGPT is useful, but is lacking in the personality department.
So I decided to make BasedGPT for fun.
Me and my friends have been getting a kick out of it, so I hope you guys do to.
Shout-out to: together.ai and NousResearch for making this one night project super easy!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: BasedGPT, CarrotWeather meets ChatGPT"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.basedgpt.chat/"}},"_tags":["story","author_GravyPouch","story_39425832","show_hn"],"author":"GravyPouch","created_at":"2024-02-19T03:00:00Z","created_at_i":1708311600,"num_comments":0,"objectID":"39425832","points":2,"story_id":39425832,"story_text":"I felt like ChatGPT is useful, but is lacking in the personality department.
So I decided to make BasedGPT for fun.
Me and my friends have been getting a kick out of it, so I hope you guys do to.
Shout-out to: together.ai and NousResearch for making this one night project super easy!","title":"Show HN: BasedGPT, CarrotWeather meets ChatGPT","updated_at":"2024-09-20T16:24:18Z","url":"https://www.basedgpt.chat/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"smohan88"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"Hey HN! \nI\u2019m building Pomi, a standalone voice agent that, with a debut launch on the Rabbit R1. Here\u2019s a 30-second video of me asking it for the weather in San Francisco: https://youtu.be/vxG5is_nNqU
I started this project because I find the R1 unusually compelling as a piece of hardware: its small, characterful, and designed specifically for interaction with an AI agent. One thing to say off the bat, this wasn't in any way motivated by frustration with Rabbit\u2019s software or an attempt to \u201cfix\u201d it. I wanted to explore a different question: how much of a custom, persistent voice-agent experience could I build directly on the device?
I had to build the UI from scratch for this to accommodate R1\u2019s 2.88-inch square display. You tap the tomato to talk, see the live transcription in a foreground overlay, and the message is sent automatically after a short pause. The response appears in the same conversation overlay. There is also an ambient lock display, conversation management, an animated home state, and an offline \u201cHey Pomi\u201d wake word using Picovoice.
The less visible part of the project was making the agent runtime behave like part of the device rather than a development process. I used Hermes by Nous Research - fantastic platform and I use it heavily outside of this. The agent gateway starts during boot, stores its state locally, and is supervised by Android\u2019s service system. Normally that runs on localhost, but I connected the R1 to my private Tailscale network and had a custom tailwind provisioned URL that only other devices (like my Mac) can access.
The repository includes the Android shell, the Hermes runtime configuration, the UI lab I used to iterate on the 240\u00d7282 interface, and scripts for building, provisioning, installing, updating, checking, and recovering the device. It targets a compatible rooted Rabbit R1 running Android 13. It does not redistribute Rabbit firmware or any user credentials.
I originally built this only for my own R1, but I cleaned up the project so other R1 owners can reproduce it: https://github.com/sidmohan0/pomi\nI\u2019ve just started collecting public feedback, so please let me know what you think. I\u2019m particularly interested in the voice interaction model, reducing the time between speaking and receiving a response, and whether the installation process is practical enough for other R1 owners. I\u2019d also love to hear what people would build with this hardware once the original software is no longer the defining constraint.
PS I used Tailscale which made things like OTA updates and accessing the Hermes Dashboard from my computer much less painful, but it isn't required!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Pomi \u2013 standalone voice agent running on the Rabbit R1"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://github.com/sidmohan0/pomi"}},"_tags":["story","author_smohan88","story_49116839","show_hn"],"author":"smohan88","created_at":"2026-07-30T22:49:03Z","created_at_i":1785451743,"num_comments":0,"objectID":"49116839","points":1,"story_id":49116839,"story_text":"Hey HN! \nI\u2019m building Pomi, a standalone voice agent that, with a debut launch on the Rabbit R1. Here\u2019s a 30-second video of me asking it for the weather in San Francisco: https://youtu.be/vxG5is_nNqU
I started this project because I find the R1 unusually compelling as a piece of hardware: its small, characterful, and designed specifically for interaction with an AI agent. One thing to say off the bat, this wasn't in any way motivated by frustration with Rabbit\u2019s software or an attempt to \u201cfix\u201d it. I wanted to explore a different question: how much of a custom, persistent voice-agent experience could I build directly on the device?
I had to build the UI from scratch for this to accommodate R1\u2019s 2.88-inch square display. You tap the tomato to talk, see the live transcription in a foreground overlay, and the message is sent automatically after a short pause. The response appears in the same conversation overlay. There is also an ambient lock display, conversation management, an animated home state, and an offline \u201cHey Pomi\u201d wake word using Picovoice.
The less visible part of the project was making the agent runtime behave like part of the device rather than a development process. I used Hermes by Nous Research - fantastic platform and I use it heavily outside of this. The agent gateway starts during boot, stores its state locally, and is supervised by Android\u2019s service system. Normally that runs on localhost, but I connected the R1 to my private Tailscale network and had a custom tailwind provisioned URL that only other devices (like my Mac) can access.
The repository includes the Android shell, the Hermes runtime configuration, the UI lab I used to iterate on the 240\u00d7282 interface, and scripts for building, provisioning, installing, updating, checking, and recovering the device. It targets a compatible rooted Rabbit R1 running Android 13. It does not redistribute Rabbit firmware or any user credentials.
I originally built this only for my own R1, but I cleaned up the project so other R1 owners can reproduce it: https://github.com/sidmohan0/pomi\nI\u2019ve just started collecting public feedback, so please let me know what you think. I\u2019m particularly interested in the voice interaction model, reducing the time between speaking and receiving a response, and whether the installation process is practical enough for other R1 owners. I\u2019d also love to hear what people would build with this hardware once the original software is no longer the defining constraint.
PS I used Tailscale which made things like OTA updates and accessing the Hermes Dashboard from my computer much less painful, but it isn't required!","title":"Show HN: Pomi \u2013 standalone voice agent running on the Rabbit R1","updated_at":"2026-07-30T22:51:15Z","url":"https://github.com/sidmohan0/pomi"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hedgehog0"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"Hi everyone,
Those who follow fine-tunes of LLMs may know that there\u2019s a company called Nous Research has been releasing a series of fine-tuned models called the Hermes, which seem to have great performance.
Since post-training is relatively cheaper than pre-training, \u201cso\u201d I also want to get into post-training and fine-tuning. Given that I'm GPU poor, with only a M4 MBP and some Tinker credits, so I was wondering if you have any advice and/or recommendations for getting into post-training? For instance, do you think this book https://www.manning.com/books/the-rlhf-book is a good place to start? If not, what\u2019s your other recommendations?
I\u2019m also currently reading \u201cHands-on LLM\u201d and \u201cBuild a LLM from scratch\u201d if that helps.
Many thanks for your time!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: Advice for getting into post-training / fine-tuning of LLMs?"}},"_tags":["story","author_hedgehog0","story_45967125","ask_hn"],"author":"hedgehog0","created_at":"2025-11-18T15:04:00Z","created_at_i":1763478240,"num_comments":0,"objectID":"45967125","points":1,"story_id":45967125,"story_text":"Hi everyone,
Those who follow fine-tunes of LLMs may know that there\u2019s a company called Nous Research has been releasing a series of fine-tuned models called the Hermes, which seem to have great performance.
Since post-training is relatively cheaper than pre-training, \u201cso\u201d I also want to get into post-training and fine-tuning. Given that I'm GPU poor, with only a M4 MBP and some Tinker credits, so I was wondering if you have any advice and/or recommendations for getting into post-training? For instance, do you think this book https://www.manning.com/books/the-rlhf-book is a good place to start? If not, what\u2019s your other recommendations?
I\u2019m also currently reading \u201cHands-on LLM\u201d and \u201cBuild a LLM from scratch\u201d if that helps.
Many thanks for your time!","title":"Ask HN: Advice for getting into post-training / fine-tuning of LLMs?","updated_at":"2026-03-05T23:02:10Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"CharlesW"},"title":{"matchLevel":"none","matchedWords":[],"value":"World_sim: LLM prompted to act as a sentient CLI universe simulator"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://worldsim.nousresearch.com/"}},"_tags":["story","author_CharlesW","story_39947713"],"author":"CharlesW","children":[39948568,39948783,39948834,39948865,39948925,39948933,39948939,39948949,39948962,39949053,39949099,39949154,39949186,39949195,39949255,39949368,39949438,39949519,39949558,39949580,39949633,39949638,39949841,39949861,39949881,39949887,39949894,39950002,39950012,39950220,39950236,39950254,39950346,39950427,39950556,39950613,39950629,39950791,39950803,39950867,39950974,39951076,39951208,39951256,39951354,39951391,39951688,39951758,39951894,39951910,39951917,39952013,39952139,39952190,39952279,39952465,39952489,39953180,39953566,39953711,39960170,39968069,39975365],"created_at":"2024-04-05T21:55:01Z","created_at_i":1712354101,"num_comments":139,"objectID":"39947713","points":297,"story_id":39947713,"title":"World_sim: LLM prompted to act as a sentient CLI universe simulator","updated_at":"2026-04-01T06:38:57Z","url":"https://worldsim.nousresearch.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"sibellavia"},"story_text":{"matchLevel":"none","matchedWords":[],"value":"Technical report: https://arxiv.org/pdf/2508.18255"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes 4"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://hermes4.nousresearch.com/"}},"_tags":["story","author_sibellavia","story_45037064"],"author":"sibellavia","children":[45069085,45069190,45069265,45069284,45069375,45069425,45069492,45069521,45069636,45069814,45069854,45070180,45070534,45070599,45070616,45070627,45071217,45071776,45071791,45072054,45072581,45072798,45074755,45074796,45075810,45077106],"created_at":"2025-08-27T08:58:38Z","created_at_i":1756285118,"num_comments":121,"objectID":"45037064","points":202,"story_id":45037064,"story_text":"Technical report: https://arxiv.org/pdf/2508.18255","title":"Hermes 4","updated_at":"2026-08-17T18:51:18Z","url":"https://hermes4.nousresearch.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"JumpCrisscross"},"title":{"matchLevel":"none","matchedWords":[],"value":"Migrate from OpenClaw"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://hermes-agent.nousresearch.com/docs/guides/migrate-from-openclaw"}},"_tags":["story","author_JumpCrisscross","story_48586005"],"author":"JumpCrisscross","children":[48586164,48586268,48586463,48586503,48586519,48586545,48586571,48586579,48586618,48586643,48586728,48586770,48586847,48587145,48587219,48587412,48588820,48589811,48591623,48592489],"created_at":"2026-06-18T14:32:13Z","created_at_i":1781793133,"num_comments":105,"objectID":"48586005","points":122,"story_id":48586005,"title":"Migrate from OpenClaw","updated_at":"2026-10-06T19:17:32Z","url":"https://hermes-agent.nousresearch.com/docs/guides/migrate-from-openclaw"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"SchwKatze"},"title":{"matchLevel":"none","matchedWords":[],"value":"DisTrO \u2013 a family of low latency distributed optimizers"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://github.com/NousResearch/DisTrO"}},"_tags":["story","author_SchwKatze","story_41371083"],"author":"SchwKatze","children":[41371084,41371891,41371992,41372305,41372670],"created_at":"2024-08-27T18:32:37Z","created_at_i":1724783557,"num_comments":26,"objectID":"41371083","points":88,"story_id":41371083,"title":"DisTrO \u2013 a family of low latency distributed optimizers","updated_at":"2024-09-20T17:42:06Z","url":"https://github.com/NousResearch/DisTrO"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"matchLevel":"none","matchedWords":[],"value":"Yarn-Mistral-7B-128k"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://huggingface.co/NousResearch/Yarn-Mistral-7b-128k"}},"_tags":["story","author_tosh","story_38233812"],"author":"tosh","children":[38233916,38234284,38234510,38234958,38235164,38235489,38235559,38236220],"created_at":"2023-11-11T19:46:16Z","created_at_i":1699731976,"num_comments":18,"objectID":"38233812","points":88,"story_id":38233812,"title":"Yarn-Mistral-7B-128k","updated_at":"2024-09-20T15:36:26Z","url":"https://huggingface.co/NousResearch/Yarn-Mistral-7b-128k"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"thunderbong"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes Desktop"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://hermes-agent.nousresearch.com/desktop"}},"_tags":["story","author_thunderbong","story_48373851"],"author":"thunderbong","children":[48374938,48377880,48382053],"created_at":"2026-06-02T18:04:32Z","created_at_i":1780423472,"num_comments":6,"objectID":"48373851","points":27,"story_id":48373851,"title":"Hermes Desktop","updated_at":"2026-08-15T12:59:54Z","url":"https://hermes-agent.nousresearch.com/desktop"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"modeless"},"title":{"matchLevel":"none","matchedWords":[],"value":"Pre-training a 15B parameter language model over the internet"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://distro.nousresearch.com/"}},"_tags":["story","author_modeless","story_42298067"],"author":"modeless","children":[42298110,42298646,42299116,42299641],"created_at":"2024-12-02T17:03:10Z","created_at_i":1733158990,"num_comments":7,"objectID":"42298067","points":26,"story_id":42298067,"title":"Pre-training a 15B parameter language model over the internet","updated_at":"2024-12-03T14:08:28Z","url":"https://distro.nousresearch.com/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes-2-Pro-Llama-3-8B"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B"}},"_tags":["story","author_tosh","story_40230591"],"author":"tosh","children":[40230817],"created_at":"2024-05-01T22:48:12Z","created_at_i":1714603692,"num_comments":0,"objectID":"40230591","points":18,"story_id":40230591,"title":"Hermes-2-Pro-Llama-3-8B","updated_at":"2024-09-20T16:58:36Z","url":"https://huggingface.co/NousResearch/Hermes-2-Pro-Llama-3-8B"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"AntiRush"},"title":{"matchLevel":"none","matchedWords":[],"value":"Measuring Thinking Efficiency in Reasoning Models: The Missing Benchmark"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://nousresearch.com/measuring-thinking-efficiency-in-reasoning-models-the-missing-benchmark/"}},"_tags":["story","author_AntiRush","story_44908035"],"author":"AntiRush","created_at":"2025-08-15T02:37:20Z","created_at_i":1755225440,"num_comments":0,"objectID":"44908035","points":9,"story_id":44908035,"title":"Measuring Thinking Efficiency in Reasoning Models: The Missing Benchmark","updated_at":"2026-03-05T22:28:32Z","url":"https://nousresearch.com/measuring-thinking-efficiency-in-reasoning-models-the-missing-benchmark/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes 3"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://twitter.com/NousResearch/status/1824131520375951454"}},"_tags":["story","author_tosh","story_41259330"],"author":"tosh","children":[41259431],"created_at":"2024-08-15T19:07:50Z","created_at_i":1723748870,"num_comments":0,"objectID":"41259330","points":9,"story_id":41259330,"title":"Hermes 3","updated_at":"2024-09-20T17:39:28Z","url":"https://twitter.com/NousResearch/status/1824131520375951454"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"meetpateltech"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes 3"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://nousresearch.com/hermes3/"}},"_tags":["story","author_meetpateltech","story_41258321"],"author":"meetpateltech","created_at":"2024-08-15T17:20:32Z","created_at_i":1723742432,"num_comments":0,"objectID":"41258321","points":7,"story_id":41258321,"title":"Hermes 3","updated_at":"2024-09-20T17:39:23Z","url":"https://nousresearch.com/hermes3/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"biraj-rocks"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes Agent can now /learn from anything"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://twitter.com/NousResearch/status/2069526242236182697"}},"_tags":["story","author_biraj-rocks","story_48653666"],"author":"biraj-rocks","children":[48654105],"created_at":"2026-06-24T00:42:03Z","created_at_i":1782261723,"num_comments":1,"objectID":"48653666","points":6,"story_id":48653666,"title":"Hermes Agent can now /learn from anything","updated_at":"2026-06-29T20:47:54Z","url":"https://twitter.com/NousResearch/status/2069526242236182697"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"pretext"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes Agent is now natively supported on Windows"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://hermes-agent.nousresearch.com/docs/user-guide/windows-native"}},"_tags":["story","author_pretext","story_48352588"],"author":"pretext","children":[48352767],"created_at":"2026-06-01T04:26:08Z","created_at_i":1780287968,"num_comments":1,"objectID":"48352588","points":5,"story_id":48352588,"title":"Hermes Agent is now natively supported on Windows","updated_at":"2026-06-18T15:47:59Z","url":"https://hermes-agent.nousresearch.com/docs/user-guide/windows-native"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes 2 \u0398"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://twitter.com/NousResearch/status/1790791623863058486"}},"_tags":["story","author_tosh","story_40370220"],"author":"tosh","created_at":"2024-05-15T17:55:46Z","created_at_i":1715795746,"num_comments":0,"objectID":"40370220","points":5,"story_id":40370220,"title":"Hermes 2 \u0398","updated_at":"2024-09-20T17:05:38Z","url":"https://twitter.com/NousResearch/status/1790791623863058486"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"matchLevel":"none","matchedWords":[],"value":"Yarn-Mistral-7B-128k"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://huggingface.co/NousResearch/Yarn-Mistral-7b-128k"}},"_tags":["story","author_tosh","story_38139477"],"author":"tosh","children":[38139538],"created_at":"2023-11-04T09:46:56Z","created_at_i":1699091216,"num_comments":2,"objectID":"38139477","points":4,"story_id":38139477,"title":"Yarn-Mistral-7B-128k","updated_at":"2024-09-20T15:35:48Z","url":"https://huggingface.co/NousResearch/Yarn-Mistral-7b-128k"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["nous"],"value":"Nous-Capybara-34B-200k"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://huggingface.co/NousResearch/Nous-Capybara-34B"}},"_tags":["story","author_tosh","story_38266768"],"author":"tosh","children":[38266830],"created_at":"2023-11-14T17:49:30Z","created_at_i":1699984170,"num_comments":1,"objectID":"38266768","points":4,"story_id":38266768,"title":"Nous-Capybara-34B-200k","updated_at":"2024-09-20T15:40:09Z","url":"https://huggingface.co/NousResearch/Nous-Capybara-34B"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"duck"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes Agent \u2013 Open-Source AI Agent That Grows with You"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://hermes-agent.nousresearch.com"}},"_tags":["story","author_duck","story_49441992"],"author":"duck","created_at":"2026-08-25T23:14:41Z","created_at_i":1787699681,"num_comments":0,"objectID":"49441992","points":4,"story_id":49441992,"title":"Hermes Agent \u2013 Open-Source AI Agent That Grows with You","updated_at":"2026-08-26T21:29:08Z","url":"https://hermes-agent.nousresearch.com"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"aurenvale"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes MoA virtual models:8% higher than Opus 4.8, 11% higher than GPT 5.5"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://twitter.com/NousResearch/status/2070610321278988385"}},"_tags":["story","author_aurenvale","story_48705599"],"author":"aurenvale","created_at":"2026-06-28T08:59:58Z","created_at_i":1782637198,"num_comments":0,"objectID":"48705599","points":4,"story_id":48705599,"title":"Hermes MoA virtual models:8% higher than Opus 4.8, 11% higher than GPT 5.5","updated_at":"2026-06-28T11:00:18Z","url":"https://twitter.com/NousResearch/status/2070610321278988385"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"gmays"},"title":{"matchLevel":"none","matchedWords":[],"value":"Lighthouse Attention"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://nousresearch.com/lighthouse-attention"}},"_tags":["story","author_gmays","story_48200742"],"author":"gmays","created_at":"2026-05-19T22:51:33Z","created_at_i":1779231093,"num_comments":0,"objectID":"48200742","points":4,"story_id":48200742,"title":"Lighthouse Attention","updated_at":"2026-05-20T03:46:00Z","url":"https://nousresearch.com/lighthouse-attention"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"talos"},"title":{"matchLevel":"none","matchedWords":[],"value":"Atropos: A rollout framework for reinforcement learning with foundation models"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://github.com/NousResearch/atropos"}},"_tags":["story","author_talos","story_43841563"],"author":"talos","created_at":"2025-04-30T05:29:37Z","created_at_i":1745990977,"num_comments":0,"objectID":"43841563","points":4,"story_id":43841563,"title":"Atropos: A rollout framework for reinforcement learning with foundation models","updated_at":"2025-04-30T17:15:06Z","url":"https://github.com/NousResearch/atropos"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"matchLevel":"none","matchedWords":[],"value":"DeepHermes Preview models: DeepHermes 24B and 3B"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://twitter.com/NousResearch/status/1900218445763088766"}},"_tags":["story","author_tosh","story_43354823"],"author":"tosh","children":[43354850],"created_at":"2025-03-13T16:22:37Z","created_at_i":1741882957,"num_comments":0,"objectID":"43354823","points":4,"story_id":43354823,"title":"DeepHermes Preview models: DeepHermes 24B and 3B","updated_at":"2025-03-20T18:17:56Z","url":"https://twitter.com/NousResearch/status/1900218445763088766"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"matchLevel":"none","matchedWords":[],"value":"Obsidian-3B"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://huggingface.co/NousResearch/Obsidian-3B-V0.5"}},"_tags":["story","author_tosh","story_38412434"],"author":"tosh","created_at":"2023-11-25T09:41:00Z","created_at_i":1700905260,"num_comments":0,"objectID":"38412434","points":4,"story_id":38412434,"title":"Obsidian-3B","updated_at":"2024-09-20T15:44:59Z","url":"https://huggingface.co/NousResearch/Obsidian-3B-V0.5"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"matchLevel":"none","matchedWords":[],"value":"Yarn-Llama-2-70B-32k"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://huggingface.co/NousResearch/Yarn-Llama-2-70b-32k"}},"_tags":["story","author_tosh","story_38351842"],"author":"tosh","created_at":"2023-11-20T18:11:13Z","created_at_i":1700503873,"num_comments":0,"objectID":"38351842","points":4,"story_id":38351842,"title":"Yarn-Llama-2-70B-32k","updated_at":"2024-09-20T15:40:50Z","url":"https://huggingface.co/NousResearch/Yarn-Llama-2-70b-32k"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"nhecker"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes Agent has 23000 open PRs and climbing"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://github.com/NousResearch/hermes-agent/pulls"}},"_tags":["story","author_nhecker","story_49424643"],"author":"nhecker","children":[49424644],"created_at":"2026-08-24T19:22:36Z","created_at_i":1787599356,"num_comments":1,"objectID":"49424643","points":3,"story_id":49424643,"title":"Hermes Agent has 23000 open PRs and climbing","updated_at":"2026-08-24T19:56:00Z","url":"https://github.com/NousResearch/hermes-agent/pulls"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"danebalia"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes Agent: The self-improving AI agent"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://github.com/NousResearch/hermes-agent"}},"_tags":["story","author_danebalia","story_47341909"],"author":"danebalia","children":[47341910],"created_at":"2026-03-11T21:09:43Z","created_at_i":1773263383,"num_comments":1,"objectID":"47341909","points":3,"story_id":47341909,"title":"Hermes Agent: The self-improving AI agent","updated_at":"2026-03-12T01:37:34Z","url":"https://github.com/NousResearch/hermes-agent"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"namenumber"},"title":{"matchLevel":"none","matchedWords":[],"value":"Democratizing AI: The Psyche Network Architecture"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://nousresearch.com/nous-psyche/"}},"_tags":["story","author_namenumber","story_43989093"],"author":"namenumber","children":[43989583],"created_at":"2025-05-14T20:52:35Z","created_at_i":1747255955,"num_comments":1,"objectID":"43989093","points":3,"story_id":43989093,"title":"Democratizing AI: The Psyche Network Architecture","updated_at":"2025-05-14T22:16:10Z","url":"https://nousresearch.com/nous-psyche/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"tosh"},"title":{"fullyHighlighted":false,"matchLevel":"partial","matchedWords":["nous"],"value":"Nous-Hermes-2-SOLAR-10.7B"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://huggingface.co/NousResearch/Nous-Hermes-2-SOLAR-10.7B"}},"_tags":["story","author_tosh","story_38843423"],"author":"tosh","children":[38843755],"created_at":"2024-01-02T16:28:39Z","created_at_i":1704212919,"num_comments":1,"objectID":"38843423","points":3,"story_id":38843423,"title":"Nous-Hermes-2-SOLAR-10.7B","updated_at":"2024-09-20T16:02:30Z","url":"https://huggingface.co/NousResearch/Nous-Hermes-2-SOLAR-10.7B"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"simonpure"},"title":{"matchLevel":"none","matchedWords":[],"value":"Hermes Agent Background Computer Use on macOS via SkyLight Private SPIs"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://hermes-agent.nousresearch.com/docs/user-guide/features/computer-use"}},"_tags":["story","author_simonpure","story_48102167"],"author":"simonpure","created_at":"2026-05-11T23:29:43Z","created_at_i":1778542183,"num_comments":0,"objectID":"48102167","points":3,"story_id":48102167,"title":"Hermes Agent Background Computer Use on macOS via SkyLight Private SPIs","updated_at":"2026-05-14T12:42:54Z","url":"https://hermes-agent.nousresearch.com/docs/user-guide/features/computer-use"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"simonpure"},"title":{"matchLevel":"none","matchedWords":[],"value":"Forge Reasoning API Beta"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://nousresearch.com/introducing-the-forge-reasoning-api-beta-and-nous-chat-an-evolution-in-llm-inference/"}},"_tags":["story","author_simonpure","story_42127801"],"author":"simonpure","created_at":"2024-11-13T17:03:00Z","created_at_i":1731517380,"num_comments":0,"objectID":"42127801","points":3,"story_id":42127801,"title":"Forge Reasoning API Beta","updated_at":"2024-11-14T08:23:31Z","url":"https://nousresearch.com/introducing-the-forge-reasoning-api-beta-and-nous-chat-an-evolution-in-llm-inference/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"anjneymidha"},"title":{"matchLevel":"none","matchedWords":[],"value":"A Preliminary Report on DisTrO [pdf]"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["nous","research"],"value":"https://github.com/NousResearch/DisTrO/blob/main/A_Preliminary_Report_on_DisTrO.pdf"}},"_tags":["story","author_anjneymidha","story_41362068"],"author":"anjneymidha","created_at":"2024-08-26T20:57:21Z","created_at_i":1724705841,"num_comments":0,"objectID":"41362068","points":3,"story_id":41362068,"title":"A Preliminary Report on DisTrO [pdf]","updated_at":"2024-09-20T17:41:13Z","url":"https://github.com/NousResearch/DisTrO/blob/main/A_Preliminary_Report_on_DisTrO.pdf"}],"hitsPerPage":50,"nbHits":119,"nbPages":3,"page":0,"params":"query=Nous+Research&tags=story&hitsPerPage=50&advancedSyntax=true&analyticsTags=backend","processingTimeMS":31,"processingTimingsMS":{"_request":{"roundTrip":17},"afterFetch":{"format":{"total":1},"merge":{"mergeLoop":{"prepareNextHit":1,"total":1},"total":1},"total":1},"fetch":{"query":7,"scanning":21,"total":29},"total":31},"query":"Nous Research","serverTimeMS":33}