By No Code MBA · Facts checked
- Provider
- Mistral
- Announcement date
- Status at last check
- Limited access or preview
- Model family
- Mistral Large
- Preceding model
- Mistral Large 3
- Capabilities
- Text, Image
What changed
Mistral Large 4 is a general-purpose multimodal model using a granular mixture-of-experts design with 1.05 trillion total parameters and a 1.6B vision encoder for image input. Mistral's docs list a 1M-token context window and support for structured outputs, function calling, document Q&A, batching, agents and built-in tools. The docs tag the release v26.10.
Access and availability
Available as a public preview on Mistral's API (La Plateforme). The docs page shows the identifier as mistral-large-4+1 in one place and mistral-large-4-0 in a link, so confirm the model ID in the console.
Official access documentationPricing
$1.36 per million input tokens, $0.14 per million cached input tokens and $4.18 per million output tokens
Official pricingWhat you can build
Compare it with your current model on representative long-context text-and-image tasks, checking answer quality, latency, and failure cases before relying on preview behavior.
Limitations
Public preview status. Mistral calls it open-weight, but the docs page shows no weights download or license yet, and no independent benchmarks have been published.