Mistral: Magistral Medium 2506

Text input Text output
Author's Description

Magistral is Mistral's first reasoning model. It is ideal for general purpose use requiring longer thought processing and better accuracy than with non-reasoning LLMs. From legal research and financial forecasting to software development and creative storytelling — this model solves multi-step challenges where transparency and precision are critical.

Key Specifications
Cost
$$$$$
Context
40K
Released
Jun 07, 2025
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Stop Presence Penalty Temperature Seed Structured Outputs Response Format Frequency Penalty Max Tokens Include Reasoning Tool Choice Top P Tools Reasoning
Features

This model supports the following features:

Tools Reasoning Structured Outputs Response Format
Performance Summary

Mistral: Magistral Medium 2506, Mistral's first reasoning model, demonstrates a strong overall performance profile, particularly excelling in reliability. It consistently performs in the top tier for speed, ranking in the 62nd percentile across benchmarks, indicating efficient processing. In terms of cost, it offers moderate pricing, positioned at the 23rd percentile. A standout feature is its exceptional reliability, achieving the 100th percentile, meaning it virtually eliminates technical failures and consistently provides usable responses. Across benchmark categories, Magistral Medium 2506 shows robust accuracy in Ethics (99.0%), Coding (92.0%), Email Classification (99.0%), and General Knowledge (98.5%), highlighting its versatility. Its Reasoning capabilities are strong at 76.0% accuracy, aligning with its description as a reasoning model. While its Instruction Following accuracy is 58.0%, it still ranks above average. Key strengths include its high accuracy in classification and knowledge-based tasks, coupled with its remarkable reliability and consistent speed. The model's ability to handle multi-step challenges with precision is evident in its strong reasoning performance. No significant weaknesses are apparent, though there is room for improvement in instruction following accuracy.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $2
Completion $5

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Mistral
Mistral | mistralai/magistral-medium-2506 40K $2 / 1M tokens $5 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Free Executions Accuracy Cost Duration
Other Models by mistralai