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,...

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

This model supports the following parameters:

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

This model supports the following features:

Structured Outputs Response Format Tools
Performance Summary

Mistral: Mixtral 8x22B Instruct demonstrates strong overall performance, particularly excelling in speed, consistently ranking in the top tier (80th percentile across 8 benchmarks). It also offers competitive pricing, positioned around the 52nd percentile. A standout feature is its exceptional reliability, achieving a 100% success rate across all benchmarks, indicating minimal technical failures. The model showcases perfect accuracy in handling hallucinations, correctly identifying fictional concepts. It performs well in General Knowledge (95.5%) and Ethics (95.0%), though these scores place it in the lower percentiles for accuracy within those categories. Its strengths are evident in Mathematics (86.0%) and Coding (83.0%), aligning with its described capabilities. However, the model exhibits weaknesses in Email Classification (80.0% accuracy, 7th percentile) and Instruction Following (50.0% accuracy, 41st percentile), suggesting areas for improvement in nuanced classification and complex multi-step directives. Reasoning also presents a challenge, with 54.0% accuracy.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $2
Completion $6
Input Cache Read $0.2

Price History

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