TLDRocket
Sign in

Voxtral

Model ● Covered in 4 stories + Follow

Voxtral is an open-source speech understanding model released by Mistral, available in 24B and 3B variants that transcribe audio and answer questions about it across 9 languages with 70ms latency. The model is priced competitively at less than half the cost of comparable services like OpenAI Whisper, starting at $0.001 per minute, and is integrated into Mistral's Le Chat platform for voice interaction capabilities. Voxtral is being used in applications like Agent Draw, which enables voice-to-drawing generation on canvas interfaces.

Updated 8 August 2026

Specifications

No specifications recorded yet.

Latest developments

Agent Draw creates and updates TLDraw canvases from voice descriptions techstackups.com · 2 months ago · 26
Deep Learning Weekly: Issue 449 Deep Learning Weekly · 5 months ago · 45
Le Chat dives deep. Mistral AI · 1 year ago · 26
Voxtral Mistral AI · 1 year ago · 18

Timeline

Month Quarter Year

July 2026

April 2026

Multiple AI companies release new model variants and agent systems in March 2026 Model release

July 2025

Relationships

Products & technology

  • Mistral develops this model · 4 sources
  • le Chat integrated with this model · 1 source

The daily briefing

Every AI story that matters, in your inbox by 8am.

TLDRocket reads all relevant sources, removes duplicate coverage, and summarises the day in two minutes. Follow companies and topics for alerts, or get the briefing in Slack. Free, no spam, unsubscribe anytime.