Ministral 3 14B 2512

Mistral· released 2 Dec 2025

mistralai/ministral-14b-2512
Input / 1M tokens
$0.2
Output / 1M tokens
$0.2
Context window262K tokens

Longest answer: 210K tokens

reads textreads imageopen weightstool callingvision

About

Ministral 3 14B 2512 is a Mistral model released on 2025-12-02. It accepts text and images and returns text, with a context window of 262,144 tokens and maximum output of 209,715 tokens. The model supports tools and vision, and its weights are listed as open. Its Hugging Face ID is mistralai/Ministral-3-14B-Instruct-2512, while its model ID is mistralai/ministral-14b-2512. Input and output each cost USD 0.2 per 1M tokens. The summary identifies it as the largest model in the Ministral 3 family and describes its capabilities as comparable to Mistral Small 3.2 24B. For teams evaluating a model with image input, tool support and open weights, these interface and pricing details help define the fit. The official page is on OpenRouter.

Who it is for

Ministral 3 14B 2512 may suit teams that need image and text inputs, tool support and open weights. It is relevant to builders comparing model interfaces and token costs for text-output tasks.

What is good

  • Open weights are listed.
  • Accepts text and image inputs.
  • Supports tools and vision.
  • Input costs USD 0.2 per 1M tokens.
  • Output costs USD 0.2 per 1M tokens.

What to know first

  • Outputs text only.
  • Maximum output is 209,715 tokens.
  • Released on 2025-12-02.

Verdict

Ministral 3 14B 2512 combines open weights, image input and tool support with equal listed input and output rates. Check its output limit and text-only output against your use case.

Details

Lab
Mistralopenrouter.ai · 3 Oct 2026
Context
262Kopenrouter.ai · 3 Oct 2026
Input price
$0.2 / 1Mopenrouter.ai · 3 Oct 2026
Output price
$0.2 / 1Mopenrouter.ai · 3 Oct 2026
Max output
209,715 tokensopenrouter.ai · 3 Oct 2026
Inputs
text, imageopenrouter.ai · 3 Oct 2026
Open weights
Yesopenrouter.ai · 3 Oct 2026

More newest models

See the list