Real-time avatars that change emotions as you talk
Hey HN, we're Ben and Caoimhe, cofounders of Anam. We build interactive avatars and just shipped our latest model, cara-4. This is the first avatar model which can naturally shift emotions and expressions during the conversation; a feature we’re calling “Director Notes”. The way it works is relatively simple: we have an LLM provide cues such as [laughter], [sad], [warm] interleaved with the speech, which we then condition our animation model with. To test it out, we commissioned a blind study from Mabyduck.com with 200 participants, 1,600 rated live interactions across six criteria. Cara-4…
In plain words
Anam has built cara-4, an interactive avatar model that displays changing emotions and expressions during conversations through a feature called "Director Notes." The system uses an LLM to generate emotion cues like [laughter] or [sad] that are interspersed with speech, then conditions an animation model to match these cues. It is designed for applications requiring lifelike avatar interactions. The model achieved top rankings in a blind study across criteria including visual quality, lip-sync accuracy, and perceived naturalness.
written from the facts on this page · September 2026
From the sources
In the maker’s words, at launch
Hey HN, we're Ben and Caoimhe, cofounders of Anam. We build interactive avatars and just shipped our latest model, cara-4. This is the first avatar model which can naturally shift emotions and expressions during the conversation; a feature we’re calling “Director Notes”. The way it works is relatively simple: we have an LLM provide cues such as [laughter], [sad], [warm] interleaved with the speech, which we then condition our animation model with. To test it out, we commissioned a blind study from Mabyduck.com with 200 participants, 1,600 rated live interactions across six criteria. Cara-4 ranked first overall and was preferred head-to-head over each competitor on overall experience, lip-sync, visual quality and “naturalness”. On latency, measured across a full week of live traffic, end of user-speech to first video frame is ~1.2s median. The avatar model's own share is just ~100ms; most of the rest is waiting on STT, LLM, TTS or various forms of buffering (an unsung latency killer). How the model works: cara-4 has a two-stage design, a diffusion transformer turns audio+text into motion embeddings (head pose, gaze, lip shape, expression), and a rendering model applies those to a reference image, so new faces works without finetuning. Why faces at all: they carry emotional signal that text and voice don't, and they're a more accessible medium. Anam started in part from Ben watching his gran struggle with her iPad and thinking there should be a face she could just talk to. If you’d like to test it for free go to anam.ai
Does the same job
all alternatives →
EVI 3: Understand and generate any voice2025 · ▲153Hume AI's new voice that truly understands emotion
More ai this month
the category →
I trained a 125M-parameter transformer to autocomplete piano performances in real time (~108 notes/sec on an iPhone 15). The idea is basically GitHub Copilot or Tabnine, except instead of prompting it with code, you prompt it by playing a few notes on a MIDI piano. The model then continues what you played, entirely on-device. The app is free if anyone wants to try it. Happy to answer questions about the model, training, Core ML, or the many things that didn't work.
AI · 17d ago · simedw.com
Astute▲585Automate your B2B brand going viral, with new media creators
AI · 18d ago · company-app.joinastute.com
Hey HN, Henry from Cactus here! We previously released Cactus Needle, a 14MB agentic LLM for tool call, device use, and structured extraction for phones, wearables, smart homes, small robots and microcontrollers. We got really great feedback here, and have now incorporated the suggestions to release Needle 2. The whole model is a single 14MB binary that runs a full session in 28MB of RAM; 45m parameters at 2bit compression. Needle hits 500 tokens/sec decode speed on a Raspberry Pi 5, sits between 400-1,500 tokens/sec on VR devices like Meta Quest 3S and Apple Vision Pro, and ranges…
AI · 27d ago · cactuscompute.com
Launched alongside, July 2026
the whole month →- IR
I might be the only SRE on Earth with his own bowling center. It's a more in-depth gig than you'd think. My family and I bought an abandoned 8-lane bowling center in the rural mid-west. In our small town there weren't many recreation options for families. You've heard of a food desert? This is an R&R desert. It had been abandoned for a good reason. The roof leaks, the electrical system was constantly surging, and my 70-year-old bowling equipment (still) doesn't work perfectly. The system that keeps your score is particularly interesting to me. It's the thing you watch during your game, but…
Life & fun · Jul 2026
- EElevators▲1,680
Life & fun · Jul 2026 · john.fun
- 1W18 Words▲1,160
Life & fun · Jul 2026 · 18words.com
- BA
Over the past few months, our team has been building more and more slidedecks using web frontend technologies with coding harnesses like Claude Code, but a common complaint is to make even small edits we need to edit the code either manually or via the harness. To avoid this loop, I ended up creating Bento, a single HTML file with everything you need in a slide tool including animations and shared editing. There's no install or cloud login, everything works offline. The default deck is around 560 KB and it doesn't need to fetch anything once you got it. Open it in a browser and then you can…
Dev tools · Jul 2026 · bento.page
- GG
A few days ago I found myself trying out GLM 5.2 and was really positively impressed. The capabilities and security I was getting from this LLM are similar to those I've gotten from models like Claude or GPT, and this really surprised me. But then I thought, "I wonder how it would work on a normal computer like mine," and above all, "I wonder if it would work without going into OOM on a computer like mine." So I started working with the help of agents to test this possibility. I started converting the model to int4, understanding MTP usage, and if possible implementing DSA for long context.…
AI · Jul 2026 · github.com
