LightweightMistral

Ministral 3 14B

Ministral 3 14B is Mistral's lightweight instruction-following model with tool calling capabilities and a 128K token context window.

Context 128K
Tier Lightweight
Tools Supported
Input from
$0.200 / 1M tokens
across 2 providers
Compare Prices

API Pricing

ProviderInput / 1MOutput / 1MSpeedTTFTUpdated
Together AI logo
Together AI
$0.200$0.200303 t/s298ms4/14/2026
OpenRouter logo
OpenRouter
$0.200$0.200303 t/s298ms4/14/2026

Prices updated daily. Last check: 4/14/2026

Model Details

General

Creator
Mistral
Family
Ministral
Tier
Lightweight
Context Window
128K
Modalities
Text

Capabilities

Tool Calling
Yes
Open Source
No
Subtypes
Chat Completion
Aliases
ministral-3-14b-instruct-2512, ministral-14b-2512

Strengths & Limitations

  • Tool calling support enables integration with external APIs and services
  • 128K token context window handles substantial document processing
  • Fast inference speed at 297.22 tokens per second output
  • Low latency with 293ms time to first token
  • Lightweight architecture reduces computational requirements
  • Instruction-following design optimized for chat and assistant tasks
  • Part of Mistral's established model ecosystem
  • Text-only modalities - no image, audio, or video input support
  • Proprietary model with no open-source weights available
  • Smaller parameter count may limit complex reasoning compared to frontier models
  • Newer model family with less extensive real-world testing than established alternatives

Key Features

•128K token context window
•Tool calling with structured output
•Chat completion interface
•Streaming response support
•Fast inference at 297+ tokens per second
•Low-latency response initiation
•API-based access
•Instruction-following optimization

About Ministral 3 14B

Ministral 3 14B is a lightweight model in Mistral's Ministral family, designed for efficient instruction-following tasks. As part of Mistral's model lineup, it occupies the lightweight tier below the company's larger frontier models, offering a balance of capability and computational efficiency. The model features a 128K token context window and supports tool calling functionality, enabling it to interact with external APIs and services. With text-only input and output capabilities, it processes at approximately 297 tokens per second with a time to first token of 293 milliseconds. The model is proprietary and available through API access rather than open-source distribution. Ministral 3 14B serves applications requiring moderate complexity reasoning and automation without the computational overhead of larger models. Its tool calling capabilities make it suitable for building agents and automated workflows, while its lightweight architecture enables cost-effective deployment for high-volume applications.

Common Use Cases

Ministral 3 14B is well-suited for applications requiring efficient automation and moderate reasoning capabilities. Its tool calling functionality makes it effective for building chatbots that need to query databases, retrieve real-time information, or interact with business systems. The 128K context window enables document analysis, customer support with conversation history, and content summarization tasks. Its lightweight architecture and fast inference make it appropriate for high-volume deployments like automated content moderation, simple coding assistance, and structured data extraction where cost efficiency is important but sophisticated reasoning is not required.

Frequently Asked Questions

How much does Ministral 3 14B cost per million tokens?

Ministral 3 14B pricing varies by provider and pricing type (standard vs batch). Check the pricing table above for current rates across all providers.

What is Ministral 3 14B best used for?

Ministral 3 14B excels at instruction-following tasks that require tool integration, such as building automated agents, processing documents within its 128K context window, and creating chatbots that need to query external systems. Its lightweight design makes it cost-effective for high-volume applications where moderate reasoning capability is sufficient.

How does Ministral 3 14B compare to other Mistral models?

Ministral 3 14B sits in the lightweight tier of Mistral's model family, offering faster inference and lower computational requirements than Mistral's larger models. While it may have less sophisticated reasoning capabilities than flagship Mistral models, it provides tool calling functionality and a substantial 128K context window for efficient automation tasks.