Mistral AI · language model
Ministral 3 14B 2512: pricing, context and benchmarks
Ministral 3 14B 2512 is a language model from Mistral AI released on December 2, 2025. It costs $0.2 per million input tokens and $0.2 per million output tokens, with a 262.1K-token context window. No independent score has been published for it so far.
| Price | 23/30 | |
|---|---|---|
| Context | 9/15 | |
| Features | 12/15 |
What you need to know about Ministral 3 14B 2512
Ministral 3 14B 2512 was released by Mistral AI (France) on December 2, 2025. It accepts text and images as input and produces text. Its context window reaches 262,144 tokens, larger than 58 % of the models in the registry, and it can write up to 209,715 tokens in one go.
It has been available for about 11 months, long enough for its prices and scores to have settled.
Ministral 3 14B 2512 pricing
On OpenRouter, Ministral 3 14B 2512 costs $0.2 per million input tokens and $0.2 per million output tokens. Counting three tokens read for every token written, its average price is $0.2 per million, more expensive than 22 % of paid models (3.8 times cheaper than the median of $0.76).
A concrete example: a task that sends 2,000 tokens (about 1,500 words) and receives a 500-token answer costs $0.0005. At 1,000 tasks a day, budget $15 a month. The price calculator redoes this with your own volumes.
Tokens already sent and cached are billed at $0.02 per million.
Performance and benchmarks
Epoch AI does not yet publish a capabilities index for Ministral 3 14B 2512. We therefore show no made-up performance rating: only price, context and features count towards its score.
Features, privacy and local use
- It supports tool calling, so it can drive an AI agent, call APIs or MCP servers.
- It can answer in an enforced JSON format (structured outputs).
Its weights are open and downloadable from Hugging Face: you can run it on your own machines without sending your data to a third party.
Other Mistral AI models
| Model | Released | Avg. price | ECI |
|---|---|---|---|
| Mistral Large 4 | Oct 6, 2026 | $2.07 | 150 |
| Mistral Medium 3.5 | Apr 30, 2026 | $3 | 141.4 |
| Mistral Small 4 | Mar 16, 2026 | $0.26 | — |
| Devstral 2 2512 | Dec 9, 2025 | $0.8 | — |
| Ministral 3 8B 2512 | Dec 2, 2025 | $0.15 | — |
| Ministral 3 3B 2512 | Dec 2, 2025 | $0.1 | — |
| Mistral Large 3 2512 | Dec 1, 2025 | $0.75 | — |
| Voxtral Small 24B 2507 | Oct 30, 2025 | $0.15 | — |
Verdict
Ministral 3 14B 2512 scores 73/100 on our grid, which only uses published figures.
Strengths
- Low price: cheaper than 78 % of paid models.
- Open weights: self-hostable, your data stays with you.
- Agent-ready: tool calling and structured JSON output.
Limitations
- No independent score published: performance cannot be verified.
Alternatives
- Qwen3 235B A22B Instruct 2507$0.21
- Gemma 4 31B$0.25
- Qwen3.5-9B$0.19
- Ternary Bonsai 2 27B$0.18
- Qwen3 VL 30B A3B Instruct$0.26
- Seed-2.0-Mini$0.18
Comparisons and tools
Learn more
Frequently asked questions
How much does Ministral 3 14B 2512 cost?
$0.2 per million input tokens and $0.2 per million output tokens on OpenRouter (checked October 25, 2026). A task with 2,000 tokens read and 500 written costs $0.0005.
How big is Ministral 3 14B 2512's context window?
262,144 tokens, roughly 197,000 words, with a maximum of 209,715 output tokens.
Can Ministral 3 14B 2512 run locally?
Yes, its weights are published.
Is Ministral 3 14B 2512 suitable for an AI agent?
Technically yes: it supports tool calling and structured outputs. Then check its cost for your volume.
What are the alternatives to Ministral 3 14B 2512?
The closest models on price and performance are Qwen3 235B A22B Instruct 2507, Gemma 4 31B, and Qwen3.5-9B.
Related articles
- Mistral AI: models, pricing and agentsOctober 25, 2026
- LLM benchmarks: how to read model rankingsOctober 24, 2026
- Running a local LLM: hardware, software, and limitationsOctober 24, 2026
- Open-source or closed models for your AI agents?October 5, 2026
Neutrality. No affiliate links. No vendor pays to be reviewed, rated or ranked. Reviews are directed by François Chollet, neutral and factual, and every data point is dated and sourced. How we score