LLM2024-07-24

Mistral Large 2

Mistral Large 2 is a high-performance 123B parameter model.

Overview

Mistral Large 2 is a 123B dense model that prioritizes efficiency and multilingual excellence. It is designed to be the best model that can fit on a single high-end node.

Unique Factor

The highest performance-to-parameter ratio for dense models in its class.

Key Capabilities

●

123B parameters

●

Multilingual

●

Coding optimized

Benchmarks

MMLU Score

84%

HumanEval (Coding)

85%

GPQA Diamond

70%

MATH Benchmark

78%

Top Use Cases

Multilingual Enterprise Chat

High-quality responses in French, German, Spanish, and English.

Example: “Explain our privacy policy to a German customer.”

Technical Specs

Context128,000 tokens

Params123B

LicenseMistral Research License

ArchTransformer

API Pricing

$2 / 1M input tokens

Output: $6 / 1M tokens

Access API

Developer

Mistral AI

European leader in high-performance open-weight AI models.

Prompt Library

Browse Coding Prompts →

📋

Previous Version

Mistral Large 1 →