If you want to try it out, you can either (1) go to https://studio.infinity.ai/try-inf2, or (2) post a comment in this thread describing a character and we\u2019ll generate a video for you and reply with a link. For example: \n\u201cMona Lisa saying \u2018what the heck are you smiling at?\u2019\u201d: https://bit.ly/3z8l1TM \n\u201cA 3D pixar-style gnome with a pointy red hat reciting the Declaration of Independence\u201d: https://bit.ly/3XzpTdS \n\u201cElon Musk singing Fly Me To The Moon by Sinatra\u201d: https://bit.ly/47jyC7C
Our tool at Infinity allows creators to type out a script with what they want their characters to say (and eventually, what they want their characters to do) and get a video out. We\u2019ve trained for about 11 GPU years (~$500k) so far and our model recently started getting good results, so we wanted to share it here. We are still actively training.
We had trouble creating videos of good characters with existing AI tools. Generative AI video models (like Runway and Luma) don\u2019t allow characters to speak. And talking avatar companies (like HeyGen and Synthesia) just do lip syncing on top of the previously recorded videos. This means you often get facial expressions and gestures that don\u2019t make sense with the audio, resulting in the \u201cuncanny\u201d look you can\u2019t quite put your finger on. See blog.
When we started Infinity, our V1 model took the lip syncing approach. In addition to mismatched gestures, this method had many limitations, including a finite library of actors (we had to fine-tune a model for each one with existing video footage) and an inability to animate imaginary characters.
To address these limitations in V2, we decided to train an end-to-end video diffusion transformer model that takes in a single image, audio, and other conditioning signals and outputs video. We believe this end-to-end approach is the best way to capture the full complexity and nuances of human motion and emotion. One drawback of our approach is that the model is slow despite using rectified flow (2-4x speed up) and a 3D VAE embedding layer (2-5x speed up).
Here are a few things the model does surprisingly well on: (1) it can handle multiple languages, (2) it has learned some physics (e.g. it generates earrings that dangle properly and infers a matching pair on the other ear), (3) it can animate diverse types of images (paintings, sculptures, etc) despite not being trained on those, and (4) it can handle singing. See blog.
Here are some failure modes of the model: (1) it cannot handle animals (only humanoid images), (2) it often inserts hands into the frame (very annoying and distracting), (3) it\u2019s not robust on cartoons, and (4) it can distort people\u2019s identities (noticeable on well-known figures). See blog.
Try the model here: https://studio.infinity.ai/try-inf2
We\u2019d love to hear what you think!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Infinity \u2013 Realistic AI characters that can speak"}},"_tags":["story","author_lcolucci","story_41467704","show_hn"],"author":"lcolucci","children":[41467807,41467870,41468005,41468049,41468051,41468101,41468124,41468144,41468190,41468248,41468254,41468257,41468305,41468319,41468370,41468377,41468458,41468494,41468506,41468515,41468520,41468548,41468578,41468612,41468620,41468629,41468801,41468942,41469022,41469064,41469101,41469151,41469195,41469224,41469301,41469423,41469531,41469544,41469546,41469557,41469907,41470041,41470055,41470103,41470214,41470270,41470279,41470414,41470529,41470615,41470713,41470790,41470824,41470829,41470844,41470876,41470929,41471045,41471052,41471118,41471397,41471464,41471886,41472116,41472362,41472405,41472669,41472823,41472828,41472839,41472864,41473049,41473069,41473406,41473458,41473613,41473614,41474098,41474653,41474702,41475566,41475811,41476361,41476749,41476801,41484083,41484285,41516335],"created_at":"2024-09-06T16:47:04Z","created_at_i":1725641224,"num_comments":294,"objectID":"41467704","points":481,"story_id":41467704,"story_text":"Hey HN, this is Lina, Andrew, and Sidney from Infinity AI (https://infinity.ai/). We've trained our own foundation video model focused on people. As far as we know, this is the first time someone has trained a video diffusion transformer that\u2019s driven by audio input. This is cool because it allows for expressive, realistic-looking characters that actually speak. Here\u2019s a blog with a bunch of examples: https://toinfinityai.github.io/v2-launch-page/
If you want to try it out, you can either (1) go to https://studio.infinity.ai/try-inf2, or (2) post a comment in this thread describing a character and we\u2019ll generate a video for you and reply with a link. For example: \n\u201cMona Lisa saying \u2018what the heck are you smiling at?\u2019\u201d: https://bit.ly/3z8l1TM \n\u201cA 3D pixar-style gnome with a pointy red hat reciting the Declaration of Independence\u201d: https://bit.ly/3XzpTdS \n\u201cElon Musk singing Fly Me To The Moon by Sinatra\u201d: https://bit.ly/47jyC7C
Our tool at Infinity allows creators to type out a script with what they want their characters to say (and eventually, what they want their characters to do) and get a video out. We\u2019ve trained for about 11 GPU years (~$500k) so far and our model recently started getting good results, so we wanted to share it here. We are still actively training.
We had trouble creating videos of good characters with existing AI tools. Generative AI video models (like Runway and Luma) don\u2019t allow characters to speak. And talking avatar companies (like HeyGen and Synthesia) just do lip syncing on top of the previously recorded videos. This means you often get facial expressions and gestures that don\u2019t make sense with the audio, resulting in the \u201cuncanny\u201d look you can\u2019t quite put your finger on. See blog.
When we started Infinity, our V1 model took the lip syncing approach. In addition to mismatched gestures, this method had many limitations, including a finite library of actors (we had to fine-tune a model for each one with existing video footage) and an inability to animate imaginary characters.
To address these limitations in V2, we decided to train an end-to-end video diffusion transformer model that takes in a single image, audio, and other conditioning signals and outputs video. We believe this end-to-end approach is the best way to capture the full complexity and nuances of human motion and emotion. One drawback of our approach is that the model is slow despite using rectified flow (2-4x speed up) and a 3D VAE embedding layer (2-5x speed up).
Here are a few things the model does surprisingly well on: (1) it can handle multiple languages, (2) it has learned some physics (e.g. it generates earrings that dangle properly and infers a matching pair on the other ear), (3) it can animate diverse types of images (paintings, sculptures, etc) despite not being trained on those, and (4) it can handle singing. See blog.
Here are some failure modes of the model: (1) it cannot handle animals (only humanoid images), (2) it often inserts hands into the frame (very annoying and distracting), (3) it\u2019s not robust on cartoons, and (4) it can distort people\u2019s identities (noticeable on well-known figures). See blog.
Try the model here: https://studio.infinity.ai/try-inf2
We\u2019d love to hear what you think!","title":"Show HN: Infinity \u2013 Realistic AI characters that can speak","updated_at":"2026-06-28T09:02:48Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jamessmithe"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"Ai generated videos are changing the whole dimension of video creation these days. Recently, I needed some video for projects and I was searching for some tools that can get the job done for me. When I found the tool, my mind was just bloen away with the videos I got from there. The quality and clarity of the videos was simply awesome. I mostly used synthesia, invideo and lunabloomai. When i saw the result, I had this question in my mind. What is the future of the freelance editors since these tools are giving such awesome results? What do you guys think?"},"title":{"matchLevel":"none","matchedWords":[],"value":"AI video generation is a threat to freelance editors"}},"_tags":["story","author_jamessmithe","story_45331837","ask_hn"],"author":"jamessmithe","children":[45333047,45333386,45336891,45339522,45341898,45344105,45344342,45345158,45345183,45345696],"created_at":"2025-09-22T11:15:09Z","created_at_i":1758539709,"num_comments":11,"objectID":"45331837","points":6,"story_id":45331837,"story_text":"Ai generated videos are changing the whole dimension of video creation these days. Recently, I needed some video for projects and I was searching for some tools that can get the job done for me. When I found the tool, my mind was just bloen away with the videos I got from there. The quality and clarity of the videos was simply awesome. I mostly used synthesia, invideo and lunabloomai. When i saw the result, I had this question in my mind. What is the future of the freelance editors since these tools are giving such awesome results? What do you guys think?","title":"AI video generation is a threat to freelance editors","updated_at":"2026-03-05T22:41:39Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"imeduardog"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"My name is Eduardo and I wanted to share I project I worked on between Dec 2020 and Dec 2024.
Mozart is an all-in-one touch display piano. It is a digital piano with a 32-in touch display that powers a music-specific software ecosystem, similar to iOS.
Back when I started the project in Dec 2020, Peloton was booming -- I loved the concept of smart-hardware and removing all friction in the experience of using a consumer electronics device. What I didn't like, was the forceful push of subscriptions on every hardware product -- both these ideas I kept in mind quite closely throughout the project. I've played the piano for over 20 years so I started exploring the possibility of doing something simialr on the piano.
I wanted to attend to the following points\n> Increasing the tablet size (beyond A4 --> multiple pages)\n> Creating a frictionless experience (no Bluetooth connecting to an tablet -- directly integrated) - similar to iOS, with core apps and potentially a future app ecosystem\n> Core apps for most important features across multiple audiences (playing, recording, learning, teaching)\n> Making markets efficient (bringing any sheet music, sound, course instantly to the piano - 1-click/touch, without all this weird management, transfering and printing)\n> Delivering something exciting, fun and new -- telling compelling story to revitalize the interest in musical instruments in younger people
Yes, initially I thought of this as a startup and went as far as a crowdfunding campaign -- which did not work out b/c I was trouble gaining attention + interest.
I also tried the accelerator and VC route -- took me a little too long to realize this, but the instruments industry is quite small ($ and units) -- rejected from YC a few times and VCs were never interested.
I self-funded the entire project -- most expenses were rent/food for the first year, then a good chunk when towards manufacturing & legal (got a patent --> which feels stupid in retrospect given how much it cost). In total, I probably invested over $40k in the project over 3y.
[Technical] There were 2 prototypes: prototype 1 was a plywood structure with the same keyboard and internal electronics -- just to show something physical to people and attract attention - prtotoype 1 was proper MDF, sanded, painted, glossed, in an attempt to replicate the designs and renders as close as possible. Hardware was an RPi4 CM4+IO-board, commodity Amazon speakers, Doepfer LM4K MIDI controller, HiFiBerry amp, and Fatar keyboard. Took a while to get it hooked up and working, but there's demos on Youtube of me playing it.
In the coming weeks, I'm going to publish the entire documentation process - conceptualization, designs, manufacturing and more -- it was a challenging yet fulfilling experience, entirely self-funded.
Happy to hear thoughts, comments, feedback about the concept and project.
As a side note -- I'm a software engineer in fintech right now, but would love to work and live in the US/Cali -- if anyone out there has an opportunities, please reach out: eduardo@garza.ca
https://mozartpianos.com\nhttps://youtube.com/@mozartpianos
* Disclaimers:
> Logo: Funny story -- I wanted a "swan" logo and I hired a guy to do one for me -- guy fleeced me and gave me an old version of the Swarovski logo -- I know nothing about women's jewlery so I only found out about it months ago. No malintent here to steal another company's logo.
> Synthesia: Full reference to an app called SeeMusic that I used to record some of the prototype/demo reels related to the guitar-hero-like notes."},"title":{"matchLevel":"none","matchedWords":[],"value":"All-in-one 32-in touch display piano (Mozart Piano)"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://www.youtube.com/channel/UCm8fVcHzGuag8Bg3YVJbblg"}},"_tags":["story","author_imeduardog","story_42496674"],"author":"imeduardog","children":[42496885],"created_at":"2024-12-23T18:49:23Z","created_at_i":1734979763,"num_comments":1,"objectID":"42496674","points":3,"story_id":42496674,"story_text":"My name is Eduardo and I wanted to share I project I worked on between Dec 2020 and Dec 2024.
Mozart is an all-in-one touch display piano. It is a digital piano with a 32-in touch display that powers a music-specific software ecosystem, similar to iOS.
Back when I started the project in Dec 2020, Peloton was booming -- I loved the concept of smart-hardware and removing all friction in the experience of using a consumer electronics device. What I didn't like, was the forceful push of subscriptions on every hardware product -- both these ideas I kept in mind quite closely throughout the project. I've played the piano for over 20 years so I started exploring the possibility of doing something simialr on the piano.
I wanted to attend to the following points\n> Increasing the tablet size (beyond A4 --> multiple pages)\n> Creating a frictionless experience (no Bluetooth connecting to an tablet -- directly integrated) - similar to iOS, with core apps and potentially a future app ecosystem\n> Core apps for most important features across multiple audiences (playing, recording, learning, teaching)\n> Making markets efficient (bringing any sheet music, sound, course instantly to the piano - 1-click/touch, without all this weird management, transfering and printing)\n> Delivering something exciting, fun and new -- telling compelling story to revitalize the interest in musical instruments in younger people
Yes, initially I thought of this as a startup and went as far as a crowdfunding campaign -- which did not work out b/c I was trouble gaining attention + interest.
I also tried the accelerator and VC route -- took me a little too long to realize this, but the instruments industry is quite small ($ and units) -- rejected from YC a few times and VCs were never interested.
I self-funded the entire project -- most expenses were rent/food for the first year, then a good chunk when towards manufacturing & legal (got a patent --> which feels stupid in retrospect given how much it cost). In total, I probably invested over $40k in the project over 3y.
[Technical] There were 2 prototypes: prototype 1 was a plywood structure with the same keyboard and internal electronics -- just to show something physical to people and attract attention - prtotoype 1 was proper MDF, sanded, painted, glossed, in an attempt to replicate the designs and renders as close as possible. Hardware was an RPi4 CM4+IO-board, commodity Amazon speakers, Doepfer LM4K MIDI controller, HiFiBerry amp, and Fatar keyboard. Took a while to get it hooked up and working, but there's demos on Youtube of me playing it.
In the coming weeks, I'm going to publish the entire documentation process - conceptualization, designs, manufacturing and more -- it was a challenging yet fulfilling experience, entirely self-funded.
Happy to hear thoughts, comments, feedback about the concept and project.
As a side note -- I'm a software engineer in fintech right now, but would love to work and live in the US/Cali -- if anyone out there has an opportunities, please reach out: eduardo@garza.ca
https://mozartpianos.com\nhttps://youtube.com/@mozartpianos
* Disclaimers:
> Logo: Funny story -- I wanted a "swan" logo and I hired a guy to do one for me -- guy fleeced me and gave me an old version of the Swarovski logo -- I know nothing about women's jewlery so I only found out about it months ago. No malintent here to steal another company's logo.
> Synthesia: Full reference to an app called SeeMusic that I used to record some of the prototype/demo reels related to the guitar-hero-like notes.","title":"All-in-one 32-in touch display piano (Mozart Piano)","updated_at":"2024-12-25T20:35:24Z","url":"https://www.youtube.com/channel/UCm8fVcHzGuag8Bg3YVJbblg"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jamessmithe"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"I was scrolling through Instagram today while I came across not less that 5 ads which are using AI generated videos for promoting their product. Most of them are using to promote their stories as they have some kind of podcast or online story telling platform. These looks good but does that somehow decrease the value of human figure in such places? These day, there are many tools like Fliki, Synthesia, Lunabloom Ai, which can do the job in just one click but what about the value of actual human touch in such videos? Isn't AI actually replacing humans?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Is AI video the future of storytelling, or just the end of stories?"}},"_tags":["story","author_jamessmithe","story_45248180","ask_hn"],"author":"jamessmithe","children":[45248848],"created_at":"2025-09-15T10:39:52Z","created_at_i":1757932792,"num_comments":2,"objectID":"45248180","points":1,"story_id":45248180,"story_text":"I was scrolling through Instagram today while I came across not less that 5 ads which are using AI generated videos for promoting their product. Most of them are using to promote their stories as they have some kind of podcast or online story telling platform. These looks good but does that somehow decrease the value of human figure in such places? These day, there are many tools like Fliki, Synthesia, Lunabloom Ai, which can do the job in just one click but what about the value of actual human touch in such videos? Isn't AI actually replacing humans?","title":"Is AI video the future of storytelling, or just the end of stories?","updated_at":"2026-03-05T22:41:29Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"hireshbrem"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"Hey HN!
I've always wanted to create videos with AI using tools like HeyGen and Synthesia, but they\u2019re just too expensive. And I know I'm not the only one who thinks this way.
So, I decided to do something about it.
Introducing Echofy!
Echofy is a simple, cheap alternative to HeyGen. Now, you can generate AI avatar videos for literally half the cost.
I'd love to hear what you think!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Simple and affordable AI video generation"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://echofyai.vercel.app"}},"_tags":["story","author_hireshbrem","story_40511538","show_hn"],"author":"hireshbrem","children":[40517944],"created_at":"2024-05-29T13:02:12Z","created_at_i":1716987732,"num_comments":1,"objectID":"40511538","points":1,"story_id":40511538,"story_text":"Hey HN!
I've always wanted to create videos with AI using tools like HeyGen and Synthesia, but they\u2019re just too expensive. And I know I'm not the only one who thinks this way.
So, I decided to do something about it.
Introducing Echofy!
Echofy is a simple, cheap alternative to HeyGen. Now, you can generate AI avatar videos for literally half the cost.
I'd love to hear what you think!","title":"Show HN: Simple and affordable AI video generation","updated_at":"2024-09-20T17:08:10Z","url":"https://echofyai.vercel.app"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"jakelazylion"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"I'm quite impressed by the video quality generated by tools like HeyGen, Synthesia where a "real" spokesperson will read through custom script in an extremely natural way.
I can't find out what kind of techniques are used, can anyone share more details or point me to research papers to learn more?"},"title":{"matchLevel":"none","matchedWords":[],"value":"Ask HN: How does AI generated spokesperson videos work?"}},"_tags":["story","author_jakelazylion","story_39075497","ask_hn"],"author":"jakelazylion","children":[39084458],"created_at":"2024-01-21T04:16:04Z","created_at_i":1705810564,"num_comments":1,"objectID":"39075497","points":1,"story_id":39075497,"story_text":"I'm quite impressed by the video quality generated by tools like HeyGen, Synthesia where a "real" spokesperson will read through custom script in an extremely natural way.
I can't find out what kind of techniques are used, can anyone share more details or point me to research papers to learn more?","title":"Ask HN: How does AI generated spokesperson videos work?","updated_at":"2024-09-20T16:16:31Z"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"chialunhera"},"story_text":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"Hi HN, we've built an API that turns text prompts into professional motion graphics animations.\nGenerative video models directly output pixels, producing a flat video file with no underlying structure and no way to edit individual elements. Hera, by contrast, uses LLMs to generate code that is then rendered into an animation. This approach enables far greater precision and control over every element, while ensuring clean edges, precise text, and transparent background support (alpha channels)
API Use Cases:
- AI Video Agents: Generating B-roll, lower thirds, and overlays to complement talking head avatars (e.g., HeyGen/Synthesia pipelines).
- Faceless Channels: Automating visual content for Reddit stories, trivia, or history channels on YouTube/TikTok.
- Data Storytelling: Programmatically turning datasets into animated charts, graphs, and map zooms.
- Marketing Automation: Scaling branded social assets, ads, and intros/outros without manual editing.
- EdTech: Automatically converting educational text or summaries into engaging visual explanations.
Input: Text prompt (e.g., "Kinetic typography intro saying 'Breaking News'") Output: Rendered video (MP4/MOV/GIF/WebM) API docs: https://docs.hera.video
We're looking for feedback on the API structure and other potential automation workflows you might need.
Try our API for free ($5 credits) by doing these steps:
1. Go to http://hera.video/join/HACKERNEWS
2. Go to https://app.hera.video/settings/developers
3. Click "Create key" to generate a new API key
Thanks!"},"title":{"matchLevel":"none","matchedWords":[],"value":"Show HN: Hera API \u2013 Text-to-Motion-Graphics Engine"},"url":{"matchLevel":"none","matchedWords":[],"value":"https://docs.hera.video/api-reference/introduction"}},"_tags":["story","author_chialunhera","story_46288213","show_hn"],"author":"chialunhera","created_at":"2025-12-16T13:24:30Z","created_at_i":1765891470,"num_comments":0,"objectID":"46288213","points":1,"story_id":46288213,"story_text":"Hi HN, we've built an API that turns text prompts into professional motion graphics animations.\nGenerative video models directly output pixels, producing a flat video file with no underlying structure and no way to edit individual elements. Hera, by contrast, uses LLMs to generate code that is then rendered into an animation. This approach enables far greater precision and control over every element, while ensuring clean edges, precise text, and transparent background support (alpha channels)
API Use Cases:
- AI Video Agents: Generating B-roll, lower thirds, and overlays to complement talking head avatars (e.g., HeyGen/Synthesia pipelines).
- Faceless Channels: Automating visual content for Reddit stories, trivia, or history channels on YouTube/TikTok.
- Data Storytelling: Programmatically turning datasets into animated charts, graphs, and map zooms.
- Marketing Automation: Scaling branded social assets, ads, and intros/outros without manual editing.
- EdTech: Automatically converting educational text or summaries into engaging visual explanations.
Input: Text prompt (e.g., "Kinetic typography intro saying 'Breaking News'") Output: Rendered video (MP4/MOV/GIF/WebM) API docs: https://docs.hera.video
We're looking for feedback on the API structure and other potential automation workflows you might need.
Try our API for free ($5 credits) by doing these steps:
1. Go to http://hera.video/join/HACKERNEWS
2. Go to https://app.hera.video/settings/developers
3. Click "Create key" to generate a new API key
Thanks!","title":"Show HN: Hera API \u2013 Text-to-Motion-Graphics Engine","updated_at":"2026-03-05T23:13:25Z","url":"https://docs.hera.video/api-reference/introduction"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"vladoh"},"title":{"matchLevel":"none","matchedWords":[],"value":"Create AI videos by simply typing in text"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/"}},"_tags":["story","author_vladoh","story_27328205"],"author":"vladoh","children":[27328665,27328673,27328747,27328758,27328759,27328767,27328772,27328803,27328846,27328849,27328883,27328923,27328973,27328994,27328995,27329010,27329091,27329172,27329314,27329418,27329423,27329438,27329452,27329470,27329620,27329658,27329676,27329700,27329734,27329744,27329887,27329962,27329976,27329992,27330053,27330105,27330166,27330411,27330448,27330461,27330563,27330612,27330690,27330726,27330969,27331026,27331273,27331822,27331828,27332383,27332530,27332750,27333063,27333280,27334361,27335994,27336022,27336213,27338651,27342250,27349795],"created_at":"2021-05-29T21:02:26Z","created_at_i":1622322146,"num_comments":110,"objectID":"27328205","points":256,"story_id":27328205,"title":"Create AI videos by simply typing in text","updated_at":"2025-01-18T09:31:07Z","url":"https://www.synthesia.io/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Keats"},"title":{"matchLevel":"none","matchedWords":[],"value":"HumanRF: High-Fidelity Neural Radiance Fields for Humans in Motion"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://synthesiaresearch.github.io/humanrf/"}},"_tags":["story","author_Keats","story_35946893"],"author":"Keats","children":[35948075,35949095,35949756,35954878],"created_at":"2023-05-15T12:06:05Z","created_at_i":1684152365,"num_comments":16,"objectID":"35946893","points":61,"story_id":35946893,"title":"HumanRF: High-Fidelity Neural Radiance Fields for Humans in Motion","updated_at":"2024-09-20T14:02:26Z","url":"https://synthesiaresearch.github.io/humanrf/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"alexvoica"},"title":{"matchLevel":"none","matchedWords":[],"value":"Automating code security review: Mythos-level capabilities at lower cost"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/post/automating-code-security-reviews-with-claude-mythos-level-capabilities"}},"_tags":["story","author_alexvoica","story_48133738"],"author":"alexvoica","created_at":"2026-05-14T11:05:46Z","created_at_i":1778756746,"num_comments":0,"objectID":"48133738","points":12,"story_id":48133738,"title":"Automating code security review: Mythos-level capabilities at lower cost","updated_at":"2026-05-16T19:04:33Z","url":"https://www.synthesia.io/post/automating-code-security-reviews-with-claude-mythos-level-capabilities"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"gbrindisi"},"title":{"matchLevel":"none","matchedWords":[],"value":"Scaling Vulnerability Management with AI: What Worked"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/post/scaling-vulnerability-management-with-ai-what-actually-worked"}},"_tags":["story","author_gbrindisi","story_47438533"],"author":"gbrindisi","children":[47438882,47439177],"created_at":"2026-03-19T12:52:15Z","created_at_i":1773924735,"num_comments":3,"objectID":"47438533","points":8,"story_id":47438533,"title":"Scaling Vulnerability Management with AI: What Worked","updated_at":"2026-03-26T11:48:18Z","url":"https://www.synthesia.io/post/scaling-vulnerability-management-with-ai-what-actually-worked"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"s1mon"},"title":{"matchLevel":"none","matchedWords":[],"value":"Synthesis.io creates training videos in 85 languages with 120 avatars using AI"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/learning-and-development"}},"_tags":["story","author_s1mon","story_35213791"],"author":"s1mon","created_at":"2023-03-18T22:29:05Z","created_at_i":1679178545,"num_comments":0,"objectID":"35213791","points":4,"story_id":35213791,"title":"Synthesis.io creates training videos in 85 languages with 120 avatars using AI","updated_at":"2024-09-20T13:34:24Z","url":"https://www.synthesia.io/learning-and-development"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"selimonder"},"title":{"matchLevel":"none","matchedWords":[],"value":"Express\u20112: Animatable, hyper-realistic, full-body avatars supporting long videos"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://synthesiaresearch.github.io/express-video/?trk=feed-detail_comments-list_comment-text"}},"_tags":["story","author_selimonder","story_44994914"],"author":"selimonder","created_at":"2025-08-23T10:31:35Z","created_at_i":1755945095,"num_comments":0,"objectID":"44994914","points":3,"story_id":44994914,"title":"Express\u20112: Animatable, hyper-realistic, full-body avatars supporting long videos","updated_at":"2026-03-05T22:33:45Z","url":"https://synthesiaresearch.github.io/express-video/?trk=feed-detail_comments-list_comment-text"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Brajeshwar"},"title":{"matchLevel":"none","matchedWords":[],"value":"An AI startup made a hyperrealistic deepfake of me that's so good it's scary"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.technologyreview.com/2024/04/25/1091772/new-generative-ai-avatar-deepfake-synthesia/"}},"_tags":["story","author_Brajeshwar","story_40200190"],"author":"Brajeshwar","created_at":"2024-04-29T16:10:15Z","created_at_i":1714407015,"num_comments":0,"objectID":"40200190","points":3,"story_id":40200190,"title":"An AI startup made a hyperrealistic deepfake of me that's so good it's scary","updated_at":"2024-09-20T16:55:00Z","url":"https://www.technologyreview.com/2024/04/25/1091772/new-generative-ai-avatar-deepfake-synthesia/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"alexvoica"},"title":{"matchLevel":"none","matchedWords":[],"value":"I made a clone of myself to provide quick facts about the company where I work"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/post/synthesia-ai-press-officer"}},"_tags":["story","author_alexvoica","story_49539333"],"author":"alexvoica","children":[49539362],"created_at":"2026-09-02T17:13:33Z","created_at_i":1788369213,"num_comments":1,"objectID":"49539333","points":2,"story_id":49539333,"title":"I made a clone of myself to provide quick facts about the company where I work","updated_at":"2026-09-02T17:21:15Z","url":"https://www.synthesia.io/post/synthesia-ai-press-officer"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"selimonder"},"title":{"matchLevel":"none","matchedWords":[],"value":"Interactive Avatars"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/features/avatars/interactive-avatars"}},"_tags":["story","author_selimonder","story_48675469"],"author":"selimonder","created_at":"2026-06-25T16:07:30Z","created_at_i":1782403650,"num_comments":0,"objectID":"48675469","points":2,"story_id":48675469,"title":"Interactive Avatars","updated_at":"2026-06-25T18:07:10Z","url":"https://www.synthesia.io/features/avatars/interactive-avatars"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"alexvoica"},"title":{"matchLevel":"none","matchedWords":[],"value":"Express-Voice: instant and accented identity voice cloning model"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://synthesiaresearch.github.io/express-voice/"}},"_tags":["story","author_alexvoica","story_44514094"],"author":"alexvoica","created_at":"2025-07-09T19:48:42Z","created_at_i":1752090522,"num_comments":0,"objectID":"44514094","points":2,"story_id":44514094,"title":"Express-Voice: instant and accented identity voice cloning model","updated_at":"2025-07-09T21:05:04Z","url":"https://synthesiaresearch.github.io/express-voice/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"marban"},"title":{"matchLevel":"none","matchedWords":[],"value":"An AI startup made a hyperrealistic deepfake of me that's so good it's scary"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.technologyreview.com/2024/04/25/1091772/new-generative-ai-avatar-deepfake-synthesia/"}},"_tags":["story","author_marban","story_40157098"],"author":"marban","created_at":"2024-04-25T13:11:38Z","created_at_i":1714050698,"num_comments":0,"objectID":"40157098","points":2,"story_id":40157098,"title":"An AI startup made a hyperrealistic deepfake of me that's so good it's scary","updated_at":"2024-09-20T16:54:05Z","url":"https://www.technologyreview.com/2024/04/25/1091772/new-generative-ai-avatar-deepfake-synthesia/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Digit-Al"},"title":{"matchLevel":"none","matchedWords":[],"value":"AI Company Releases Deepfakes into the Wild. Can It Control Them?"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.wired.co.uk/article/synthesia-ai-deepfakes-it-control-riparbelli"}},"_tags":["story","author_Digit-Al","story_37198957"],"author":"Digit-Al","created_at":"2023-08-20T13:30:26Z","created_at_i":1692538226,"num_comments":0,"objectID":"37198957","points":2,"story_id":37198957,"title":"AI Company Releases Deepfakes into the Wild. Can It Control Them?","updated_at":"2024-09-20T14:55:36Z","url":"https://www.wired.co.uk/article/synthesia-ai-deepfakes-it-control-riparbelli"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"RobSpectre"},"title":{"matchLevel":"none","matchedWords":[],"value":"Personalized Holiday Greetings from GAN-Powered Santa Claus"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/santa"}},"_tags":["story","author_RobSpectre","story_25294907"],"author":"RobSpectre","created_at":"2020-12-03T21:05:00Z","created_at_i":1607029500,"num_comments":0,"objectID":"25294907","points":2,"story_id":25294907,"title":"Personalized Holiday Greetings from GAN-Powered Santa Claus","updated_at":"2023-09-07T08:07:31Z","url":"https://www.synthesia.io/santa"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Stjerrild"},"title":{"matchLevel":"none","matchedWords":[],"value":"AI Santa"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/santa"}},"_tags":["story","author_Stjerrild","story_25287707"],"author":"Stjerrild","created_at":"2020-12-03T11:00:53Z","created_at_i":1606993253,"num_comments":0,"objectID":"25287707","points":2,"story_id":25287707,"title":"AI Santa","updated_at":"2024-09-20T07:30:55Z","url":"https://www.synthesia.io/santa"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Stjerrild"},"title":{"matchLevel":"none","matchedWords":[],"value":"Worlds first text to video platform"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/request-demo"}},"_tags":["story","author_Stjerrild","story_24430517"],"author":"Stjerrild","children":[24431288],"created_at":"2020-09-10T09:35:29Z","created_at_i":1599730529,"num_comments":0,"objectID":"24430517","points":2,"story_id":24430517,"title":"Worlds first text to video platform","updated_at":"2024-09-20T06:55:38Z","url":"https://www.synthesia.io/request-demo"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"Stjerrild"},"title":{"matchLevel":"none","matchedWords":[],"value":"AI dubbing 12x more engaging than normal dubbing"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/blog-post/synthesia-translate-improves-video-engagement-by-up-to-1200"}},"_tags":["story","author_Stjerrild","story_22042894"],"author":"Stjerrild","created_at":"2020-01-14T08:31:46Z","created_at_i":1578990706,"num_comments":0,"objectID":"22042894","points":2,"story_id":22042894,"title":"AI dubbing 12x more engaging than normal dubbing","updated_at":"2024-09-20T05:29:40Z","url":"https://www.synthesia.io/blog-post/synthesia-translate-improves-video-engagement-by-up-to-1200"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"rdougl"},"title":{"matchLevel":"none","matchedWords":[],"value":"Embeddable Interactive Avatars"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/features/avatars/interactive-avatars"}},"_tags":["story","author_rdougl","story_49738101"],"author":"rdougl","children":[49738102],"created_at":"2026-09-17T08:53:01Z","created_at_i":1789635181,"num_comments":0,"objectID":"49738101","points":1,"story_id":49738101,"title":"Embeddable Interactive Avatars","updated_at":"2026-09-17T08:53:59Z","url":"https://www.synthesia.io/features/avatars/interactive-avatars"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"selimonder"},"title":{"matchLevel":"none","matchedWords":[],"value":"Send a Free Santa Video"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/santa"}},"_tags":["story","author_selimonder","story_46211726"],"author":"selimonder","created_at":"2025-12-09T22:39:45Z","created_at_i":1765319985,"num_comments":0,"objectID":"46211726","points":1,"story_id":46211726,"title":"Send a Free Santa Video","updated_at":"2026-03-05T23:07:49Z","url":"https://www.synthesia.io/santa"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"belter"},"title":{"matchLevel":"none","matchedWords":[],"value":"An AI startup made a hyperrealistic deepfake of me that's so good it's scary"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.technologyreview.com/2024/04/25/1091772/new-generative-ai-avatar-deepfake-synthesia/"}},"_tags":["story","author_belter","story_40614035"],"author":"belter","created_at":"2024-06-07T23:42:28Z","created_at_i":1717803748,"num_comments":0,"objectID":"40614035","points":1,"story_id":40614035,"title":"An AI startup made a hyperrealistic deepfake of me that's so good it's scary","updated_at":"2024-09-20T17:11:29Z","url":"https://www.technologyreview.com/2024/04/25/1091772/new-generative-ai-avatar-deepfake-synthesia/"},{"_highlightResult":{"author":{"matchLevel":"none","matchedWords":[],"value":"teleforce"},"title":{"matchLevel":"none","matchedWords":[],"value":"50 Video Statistics You Can't Ignore in 2024"},"url":{"fullyHighlighted":false,"matchLevel":"full","matchedWords":["synthesia"],"value":"https://www.synthesia.io/post/video-statistics"}},"_tags":["story","author_teleforce","story_40510934"],"author":"teleforce","created_at":"2024-05-29T11:54:30Z","created_at_i":1716983670,"num_comments":0,"objectID":"40510934","points":1,"story_id":40510934,"title":"50 Video Statistics You Can't Ignore in 2024","updated_at":"2024-09-20T17:08:02Z","url":"https://www.synthesia.io/post/video-statistics"}],"hitsPerPage":50,"nbHits":5765,"nbPages":20,"page":0,"params":"query=Synthesia&tags=story&hitsPerPage=50&advancedSyntax=true&analyticsTags=backend","processingTimeMS":12,"processingTimingsMS":{"_request":{"roundTrip":18},"afterFetch":{"format":{"highlighting":1,"total":1},"merge":{"mergeLoop":{"prepareNextHit":1,"total":1},"total":1},"total":1},"fetch":{"query":6,"scanning":2,"total":9},"total":12},"query":"Synthesia","serverTimeMS":14}