EN
Open-weight models

Ministral 3 8B 2512, the Mistral AI model in numbers

Ministral 3 8B 2512 in numbers: price per token, context window, independent benchmark scores and comparable models, based on public data checked on October 22, 2026.

Data as of October 22, 2026

Key figures

DeveloperMistral AI
ReleasedDecember 2, 2025
Context window262,144 tokens (≈ 393 pages)
Maximum output209,715 tokens
Input price$0.15 per million tokens (median of 8 providers)
Output price$0.15 per million tokens
Open weightsYes
Reasoning modeNo
Tool callingYes
Structured output (JSON)Yes
Accepted inputstext, image
Providers offering it8

Overview

Ministral 3 8B 2512 is a language model from Mistral AI, released on December 2, 2025. That makes it 10 months old as of October 22, 2026. Its weights are published: you can download it and run it on your own servers, or use one of the 8 API providers that list it on models.dev. It has no declared reasoning mode: it answers directly, without an extra billed thinking step.

We file it under open-weight models (68 models in this category on the site).

What do independent benchmarks say?

Epoch AI publishes a single result for Ministral 3 8B 2512 on our basket's benchmarks: too few to compute a reliable capability index.

  • GPQA Diamond (PhD-level science questions): 27.1% — the best tracked model, Gemini 4 Argon, reaches 96.4%

What it costs

Ministral 3 8B 2512 costs $0.15 per million input tokens and $0.15 per million output tokens (median of 8 providers, checked on October 22, 2026).

To put it in perspective: reading 20,000 tokens (about thirty pages) and writing 2,000 costs roughly $0.003; a thousand of those tasks, $3.3.

That is 6.7 times cheaper than the median output price of open-weight models ($1).

To estimate your own budget, use the agent cost calculator.

Context and capabilities

Its context window reaches 262,144 tokens, roughly 393 pages of text at 500 words per page: enough for a long report or several contracts at once. A single answer can run to 209,715 tokens.

It accepts as input: text and image.

It can call tools (functions, APIs) and return structured JSON output, which makes it a possible foundation for an AI agent.

The weights are published on Hugging Face (mistralai/Ministral-3-8B-Instruct-2512). See also our list of open-source models.

What is it good for?

Given its category, Ministral 3 8B 2512 is mainly aimed at projects where you want control over your data or hosting, or want to adapt the model to your field.

Its very low cost (under half a cent for our sample task) means it can be called thousands of times a day.

Before adopting it, test it on your own examples: benchmarks measure standard tasks, not your job.

What could replace it?

In the same category, the models closest to Ministral 3 8B 2512 are:

Independent benchmark scores

BenchmarkWhat it measuresScore
GPQA DiamondPhD-level science questions27.1%

Similar models

Other Mistral AI models

Frequently asked questions

How much does Ministral 3 8B 2512 cost?

As of October 22, 2026, Ministral 3 8B 2512 costs $0.15 per million input tokens and $0.15 per million output tokens (median of 8 providers). A task that reads 20,000 tokens and writes 2,000 costs about $0.003.

What is the context window of Ministral 3 8B 2512?

262,144 tokens, roughly 393 pages of text. An answer can reach 209,715 tokens.

Is Ministral 3 8B 2512 open source?

Its weights are published: you can install it on your own servers. Check the license before commercial use.

When was Ministral 3 8B 2512 released?

Ministral 3 8B 2512 was released on December 2, 2025, according to models.dev.

Can you build an agent with Ministral 3 8B 2512?

Yes: it can call tools and return structured JSON, the two basic building blocks of an agent.

Data sources

Data as of October 22, 2026

Report an error

Does a figure, a date or a fact look wrong? Tell us: every report is checked.