Vision Agents

@visionagents_ai

Open-source framework for building real-time voice and video AI applications ⚡visionagents.ai ⚡️

The Future

github.com/GetStream/Visi…

Replies Reposts Media Articles

Vision Agents
@visionagents_ai
Sep 2
@inworld_ai AI shipped Realtime TTS-2 today, plus TTS-2 Flash at 25ms time to first byte. For voice agents, that's the number that matters. Stream is a launch partner, and TTS-2 is already wired into Vision Agents. Swap it in with one line 👉 visionagents.ai/integrations/r…
Inworld AI
@inworld
Sep 2
Realtime TTS-2 is GA today, now the #1 model on Artificial Analysis and fastest TTS in the world of its class. The research preview got far more usage than expected. That experience went back into research and today's GA model represents the biggest leap in Inworld's history.
00:00
Vision Agents
@visionagents_ai
Aug 26
Our friends over at @googledevs gave us an early look into their new Gemini Transcribe 3.5 models. We found them to be some of the best real-world transcription models allowing extremely accurate verbatim transcription or a SMART mode filtering out filler words and unneeded
00:00
1
Vision Agents
@visionagents_ai
Aug 25
You can use LLMs for a LOT of things. "Should you?" is an entirely different question. Our very own @stefanjblos built an app with Vision Agents to help him know things he has no business knowing in an interview (look up: cheating). You can also use it for a sales assistant or
00:00
Vision Agents
@visionagents_ai
Jul 31
Model releases this week have been CRAZY, but we think the release of Grok Voice Think Fast 2.0 (handy name) went under the radar. It's an incredibly fast, accurate, and expressive voice model that is fun to talk to. Although we're not sure it really helped @stefanjblos...
00:00
3
Vision Agents
@visionagents_ai
Jul 30
We've already had support for their previous models, but have added @FishAudio 's new S2.1 Pro support - for good reason! Just see how natural it sounds and how expressive it is (while not being responsible for insulting our resident AI investigator @stefanjblos)!
Stefan
@stefanjblos
Jul 30
✅Being insulted by an AI for my sheer humanness - CHECK! Wanted to test @FishAudio 's new S2.1 Pro models and they're really fantastic! Expressiveness, latency and emotions have actually blown me away. See the video for the full example!
00:00