[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]

![ai_for_success Avatar](https://lunarcrush.com/gi/w:24/cr:twitter::1618975370488999936.png) AshutoshShrivastava [@ai_for_success](/creator/twitter/ai_for_success) on x 64K followers
Created: 2025-07-15 14:50:17 UTC

🚨Mistral just launched Voxtral, the world’s most powerful open-source speech recognition models, surpassing Whisper large v3, GPT 4o mini, Gemini XXX Flash, and ElevenLabs Scribe.
- Models: 3B and 24B, available via API, Le Chat, and Hugging Face
- Handles up to XX to XX minutes of audio (32k token context)
- Built-in Q&A and summarization from audio
- Multilingual support
- Supports function calling from voice commands

![](https://pbs.twimg.com/media/Gv6Ah1uWkAAy55t.jpg)

XXXXXX engagements

![Engagements Line Chart](https://lunarcrush.com/gi/w:600/p:tweet::1945133824016535670/c:line.svg)

**Related Topics**
[blog](/topic/blog)
[token](/topic/token)
[le](/topic/le)
[gpt](/topic/gpt)
[the worlds](/topic/the-worlds)
[just launched](/topic/just-launched)

[Post Link](https://x.com/ai_for_success/status/1945133824016535670)