Mistral: Ministral 8B

Text input Text output
Author's Description

Ministral 8B is an 8B parameter model featuring a unique interleaved sliding-window attention pattern for faster, memory-efficient inference. Designed for edge use cases, it supports up to 128k context length and excels in knowledge and reasoning tasks. It outperforms peers in the sub-10B category, making it perfect for low-latency, privacy-first applications.

Key Specifications
Cost
$$
Context
128K
Parameters
8B
Released
Oct 16, 2024
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

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

This model supports the following features:

Tools Response Format Structured Outputs
Performance Summary

Mistral's Ministral 8B model demonstrates strong performance in several key areas, particularly excelling in efficiency and reliability. It performs among the fastest models, ranking in the 77th percentile for speed, and consistently offers highly competitive pricing, placing in the 88th percentile. The model exhibits exceptional reliability with a 100% success rate across all benchmarks, indicating robust technical stability. In terms of specific benchmarks, Ministral 8B shows a notable strength in acknowledging uncertainty, achieving 96.0% accuracy in Hallucinations (Baseline) by correctly identifying fictional concepts. It also performs well in Email Classification (96.0% accuracy) and Ethics (96.5% accuracy), though its percentile rankings in these areas are moderate. Key weaknesses are apparent in complex Reasoning (20.0% accuracy, 11th percentile) and Instruction Following (40.4% accuracy, 39th percentile), suggesting limitations in handling highly intricate tasks. General Knowledge, Mathematics, and Coding benchmarks show moderate accuracy, indicating room for improvement in these domains. Overall, Ministral 8B is a highly reliable and cost-effective option, particularly suited for applications where speed, cost-efficiency, and robust operation are paramount, especially in edge use cases.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.1
Completion $0.1

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
Mistral
Mistral | mistralai/ministral-8b 128K $0.1 / 1M tokens $0.1 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Strategy Free Executions Accuracy Cost Duration
Other Models by mistralai