Question 1

What is Ministral 3 14B Instruct?

Accepted Answer

Ministral 3 14B Instruct is a model from Mistral, released on Dec 2, 2025. Ministral 3 14B Instruct is a Mistral model focused on efficient inference for chat, coding, and structured generation tasks. Ministral 3 14B Instruct is a Mistral model focused on efficient inference for chat, coding, and structured generation tasks. Up to 262K context window.

Question 2

How much does Ministral 3 14B Instruct cost?

Accepted Answer

On the Respan gateway, input starts at $0.20 per 1M tokens on TogetherAI, and output starts at $0.20 per 1M tokens on TogetherAI, after gateway discounts.

Question 3

Which providers serve Ministral 3 14B Instruct on Respan?

Accepted Answer

Ministral 3 14B Instruct is served by TogetherAI. TogetherAI requires your own provider API key (BYOK).

Question 4

What is Ministral 3 14B Instruct's context window?

Accepted Answer

Ministral 3 14B Instruct accepts up to 262K tokens of context.

Question 5

How do I use Ministral 3 14B Instruct with the Respan gateway?

Accepted Answer

Point any OpenAI-compatible SDK at https://api.respan.ai/api/ with your Respan API key and set the model to "ministral-3-14b-instruct". To set it up with a coding agent, copy the agent setup prompt at the top of this page.

Question 6

How does Ministral 3 14B Instruct score on benchmarks?

Accepted Answer

Ministral 3 14B Instruct has 3 published benchmark scores, including LiveCodeBench 64.6%, GPQA Diamond 71.2%, and AIME 2025 85%. Each score links to where it was published; the data is collected by llmmetric.com.

Ministral 3 14B Instruct

Providers

Pricing

Cost by volume

Benchmarks

Certified where it counts.

Coding1

Reasoning1

Math1

Apps