Compare Mistral Large 2 vs. Mistral Large

Pricing About Us Blog Docs

Compare

Understand and compare Mistral Large 2 vs. Mistral Large

Overview

Mistral Large 2 was released 5 months after Mistral Large.

	Mistral Large 2	Mistral Large
Provider The entity that provides this model.	Mistral	Mistral
Input Context Window The number of tokens supported by the input context window.	128K tokens	32K tokens
Maximum Output Tokens The number of tokens that can be generated by the model in a single request.	8,192 tokens	4,096 tokens
Release Date When the model was first released.	July 24th, 2024 4 months ago 2024-07-24	February 26th, 2024 9 months ago 2024-02-26
Knowledge Cutoff Limit on the knowledge base used by the model.	Unknown	Unknown
Open Source	Yes	Yes
API Providers The providers that offer this model. (This is not an exhaustive list.)	Azure AI, AWS Bedrock, Google Cloud Vertex AI Model Garden, Snowflake Cortex	Azure AI, AWS Bedrock, Google Cloud Vertex AI Model Garden, Snowflake Cortex

Pricing

Mistral Large 2 is roughly 2.7x cheaper compared to Mistral Large for input tokens and roughly 12.5% more expensive for output tokens.

	Mistral Large 2	Mistral Large
Input Cost of input data provided to the model.	$3.00 per million tokens	$8.00 per million tokens
Output Cost of output tokens generated by the model.	$9.00 per million tokens	$8.00 per million tokens

Benchmarks

Compare relevant benchmarks between Mistral Large 2 and Mistral Large.

	Mistral Large 2	Mistral Large
MMLU Evaluating LLM knowledge acquisition in zero-shot and few-shot settings.	84.0 (5-shot) Source	81.2 (5-shot) Source
MMMU A wide ranging multi-discipline and multimodal benchmark.	Benchmark not available.	Benchmark not available.
HellaSwag A challenging sentence completion benchmark.	Benchmark not available.	89.2 (10-shot) Source
GSM8K Grade-school math problems benchmark.	Benchmark not available.	Benchmark not available.
HumanEval A benchmark to measure functional correctness for synthesizing programs from docstrings.	Benchmark not available.	Benchmark not available.
MATH Benchmark performance on Math problems ranging across 5 levels of difficulty and 7 sub-disciplines.	Benchmark not available.	Benchmark not available.