Mistral: Mixtral 8x22B Instruct

Text input Text output
Author's Description

Mistral's official instruct fine-tuned version of [Mixtral 8x22B](/models/mistralai/mixtral-8x22b). It uses 39B active parameters out of 141B, offering unparalleled cost efficiency for its size. Its strengths include: - strong math, coding, and reasoning - large context length (64k) - fluency in English, French, Italian, German, and Spanish See benchmarks on the launch announcement [here](https://mistral.ai/news/mixtral-8x22b/). #moe

Key Specifications
Cost
$$$$
Context
65K
Parameters
22B
Released
Apr 16, 2024
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Stop Presence Penalty Logit Bias Temperature Structured Outputs Response Format Frequency Penalty Max Tokens Tool Choice Top P Tools Logprobs Top Logprobs
Features

This model supports the following features:

Tools Structured Outputs Response Format
Performance Summary

Mistral: Mixtral 8x22B Instruct demonstrates strong overall performance, particularly excelling in reliability, where it consistently provides usable responses with minimal technical failures, ranking in the 100th percentile. The model performs among the fastest models, with its speed ranking in the 76th percentile across benchmarks. It also offers competitive pricing, positioned in the 47th percentile. In terms of specific benchmarks, the model shows a notable strength in General Knowledge and Ethics, achieving high accuracy rates of 95.5% and 95.0% respectively, although its percentile rankings in these areas are moderate. Its Coding performance is solid at 83.0% accuracy, placing it in the 61st percentile. However, the model exhibits weaknesses in Email Classification, where its 80.0% accuracy places it in the 10th percentile, indicating a need for improvement in this specific classification task. Reasoning and Instruction Following also show room for improvement, with accuracy rates of 47.5% and 50.0% respectively. Despite some lower accuracy scores, the model's high reliability ensures consistent output, making it a dependable choice for various applications. Its large context length and multilingual fluency are key advantages.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.9
Completion $0.9

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Fireworks
Fireworks | mistralai/mixtral-8x22b-instruct 65K $0.9 / 1M tokens $0.9 / 1M tokens
Mistral
Mistral | mistralai/mixtral-8x22b-instruct 65K $2 / 1M tokens $6 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Free Executions Accuracy Cost Duration
Other Models by mistralai