Model registry / Mistral AI / Mistral Large / Mistral Large 4
Mistral Large 4
Mistral's 1T-parameter multimodal MoE for reasoning, coding, and agentic work
Line
Mistral Large
Weights
API only
Released
2026-10-06
Context
524K
Input
$0.68 / 1M
Output
$2.09 / 1M
Max output
262K
Coverage
not covered yet
Overview
Mistral Large 4 sits in the Mistral Large line at Mistral AI, released OCT 6, 2026. It accepts image, text and returns text, with a 524K-token window. The cheapest host that serves it charges $0.68 / $2.09 per Mtok.
Specs
Released2026-10-06
LineMistral AI · Mistral Large
WeightsAPI only
Context524K tokens
Max output262K tokens
Inputimage, text
Outputtext
ReasoningYes
Tool callingYes
Knowledge cutoffnot stated
API idmistral-large-4
Pricing
Input$0.68 / 1M tokens
Cached input$0.07 / 1M tokens
Output$2.09 / 1M tokens
First version of its line in the registry.
How to access
ProviderModel id
Eden AImistral-large-4
Mistralmistral-large-4
Requestymistral-large-4
Vercel AI Gatewaymistral-large-4
4 hosts serve this model at the price above · the maker's documentation
Mistral Large: every version
VersionReleasedContextInput
Mistral Large 4Current2026-10-06524K$0.68
The registry keeps the six most recent versions of a line. Older ones stay in the file and can be surfaced without a rebuild.
ProprietaryReasoning524K context