What does real-time transcription look like in a live setting? 🎙 We used Voxtral Transcribe 3 at AI Engineer Paris to capture every main-stage talk in real time. We boosted accuracy with context biasing, feeding the model speaker names and technical terms from the conference agenda..

On First Partial Transcription, Gemini 3.5 Transcribe Live achieves 5.8% WER at 0.25s after detected end of speech. GPT Live Transcribe achieves 6.3% WER at 0.26s, making Gemini slightly faster and more accurate at this measurement point.

New leaderboard for audio transcription just launched and our apache 2.0 Cohere-Transcribe is at the top. This eval didn't exist when we trained the model, so its nice to see us do so well on it.

5 GITHUB REPOS FOR AI TRANSCRIPTION: 1. WhisperX 2. Faster Whisper 3. Whisper 4. Whisper.cpp 5. Pyannote Audio

I took the full Tibo interview (official transcription) and gave it to GPT 5.6, then asked them to give an independent conclusion on what might be happening with the proposed merge, based off the things people are concerned about. Result: My entirely independent reading is that Tibo is describing a unified ChatGPT experience that hides more of the machinery underneath it, rather than announcing that everything is literally becoming one mode: the clearest near-term intention is to remove the awkward choice between ordinary Chat and more capable agentic experiences such as Work/Codex, while deliberately preserving the fact that Chat is “faster and more pleasant”; longer term, Dots’ capabilities are intended to come into ChatGPT, not replace it. He plainly dislikes configuration fatigue—including model pickers, reasoning effort and agent/mode choices—so I think increased automatic routing is a very reasonable prediction, but the interview does not establish that ChatGPT’s model picker will be removed, when that might happen, whether optional manual selection will remain, or that Chat/Work/Codex will share one usage pool. My best interpretation is therefore: one conversational front door, increasingly able to call the right specialist capability automatically, with much less need for the user to understand what is happening underneath; anything more specific about model access or quotas is currently extrapolation rather than something Tibo announced.

Deleted the circular financing comment from Hock Tan because Quartr's transcription picked it up wrong. I still think he could've worded it better but he said: "that is not circular finance where we are using financing to create demand. Demand is there."

What happens when speech, transcription, and coding models work together? This prototype demo, built using a VS Code fork, showcases how MAI-Transcribe, MAI-Voice, and MAI-Code-1-Flash can work together in a unified workflow to transform spoken instructions into working code.

Grok Voice Transcribe 2.0 just launched It's already #1# in overall transcription accuracy at 97.4%, ahead of ElevenLabs, Gemini, and Deepgram And it starts at just $0.10/hour Highest accuracy at the lowest price

Cohere Transcribe is now available on @superwhisper 💬 Push to talk and get your transcription back almost instantly. With Superwhisper, you can use Transcribe offline, integrate it into your favorite apps, and recall specialist vocabulary.