AI news · October 8, 2026
Mistral Large 4 pairs a one-million-token window with a low launch price
Mistral Large 4 offers a 1M-token context, multimodal input, and launch pricing of $0.68 per million input tokens. An open-weight release is planned for the end of October.
- total/active parameters
- 1.05T/52B
- token context window
- 1M
- input/output price per million tokens
- $0.68/$2.09
Mistral released Large 4 in public preview on October 6. Its model documentation lists about 1.05 trillion total parameters and 52 billion active parameters, with a one-million-token context window and a 1.6 billion parameter vision encoder. The model accepts text and images, supports function calling and structured output, and is designed for agent workflows. Mistral is offering launch pricing of $0.68 per million input tokens and $2.09 per million output tokens, with a lower cached-input price.
The long context makes this interesting for large codebases, document sets, and multi-step research, but a large window is not a free substitute for retrieval and careful prompting. The public preview is available through Mistral's developer surface, while the company says open weights are planned for the end of October. Builders should measure answer quality and context selection on their own data rather than assuming that more context will produce better decisions.
What you can do with it
Try the public preview through the Mistral model documentation with one long document task and one tool-calling task. Track input size, selected evidence, latency, output errors, and cost. Keep a smaller context baseline so the one-million-token window earns its place.
Our take
Large 4 gives Mistral a credible long-context option at a price that is easy to test. The planned open-weight release could matter more than the preview if it runs well on private infrastructure. Until then, the right decision is a short benchmark, not a wholesale model switch.
Source: Mistral AI ↗ — Made With Models writes the brief; the reporting is theirs.