ByteBuddy

English

简体中文

English

简体中文

深色模式

Mistral ​

Mistral AI provides high-performance open-source language models with excellent reasoning capabilities and efficiency.

Supported Models ​

Latest Models ​

  • mistral-large-latest - Latest large model (123B parameters)
  • mistral-medium-latest - Medium model
  • mistral-small-latest - Small model

Open-Source Models ​

  • open-mistral-7b - Open-source 7B model
  • open-mixtral-8x7b - Open-source MoE 8x7B model
  • open-mixtral-8x22b - Open-source MoE 8x22B model

Code Models ​

  • codestral-latest - Code-specialized model
  • codestral-2405 - Specific version of code model

Configuration ​

Basic Configuration ​

Configure in config.yaml or ~/.bytebuddy/config.yaml:

yaml
models:
  - name: "mistral-large"
    provider: "mistral"
    model: "mistral-large-latest"
    apiKey: "${MISTRAL_API_KEY}"
    roles: ["chat", "edit"]
    defaultCompletionOptions:
      temperature: 0.7
      maxTokens: 8192

Code Generation Configuration ​

yaml
models:
  - name: "codestral"
    provider: "mistral"
    model: "codestral-latest"
    apiKey: "${MISTRAL_API_KEY}"
    roles: ["edit", "apply"]
    defaultCompletionOptions:
      temperature: 0.2
      maxTokens: 4096

Multi-Model Configuration ​

yaml
models:
  - name: "mistral-large"
    provider: "mistral"
    model: "mistral-large-latest"
    apiKey: "${MISTRAL_API_KEY}"
    roles: ["chat"]
    defaultCompletionOptions:
      temperature: 0.7
      maxTokens: 8192

  - name: "mistral-small"
    provider: "mistral"
    model: "mistral-small-latest"
    apiKey: "${MISTRAL_API_KEY}"
    roles: ["chat", "autocomplete"]
    defaultCompletionOptions:
      temperature: 0.5
      maxTokens: 4096

  - name: "codestral"
    provider: "mistral"
    model: "codestral-latest"
    apiKey: "${MISTRAL_API_KEY}"
    roles: ["edit", "apply"]
    defaultCompletionOptions:
      temperature: 0.2
      maxTokens: 4096

Configuration Fields ​

Required Fields ​

  • name: Unique identifier for the model configuration
  • provider: Set to "mistral"
  • model: Model name
  • apiKey: Mistral API key

Optional Fields ​

  • roles: Model roles [chat, edit, apply, autocomplete]
  • defaultCompletionOptions:
    • temperature: Control randomness (0-1)
    • maxTokens: Maximum tokens
    • topP: Nucleus sampling parameter
    • presencePenalty: Presence penalty (-2 to 2)
    • frequencyPenalty: Frequency penalty (-2 to 2)
    • stopSequences: Stop sequences

Environment Variables ​

bash
# ~/.bashrc or ~/.zshrc
export MISTRAL_API_KEY="your-mistral-api-key"

Getting API Key ​

  1. Visit Mistral AI Console
  2. Register and log in to account
  3. Create new key in API Keys page
  4. Save the key to environment variable

Use Case Configurations ​

High-Quality Chat ​

yaml
models:
  - name: "quality-chat"
    provider: "mistral"
    model: "mistral-large-latest"
    apiKey: "${MISTRAL_API_KEY}"
    roles: ["chat"]
    defaultCompletionOptions:
      temperature: 0.7
      maxTokens: 8192

Code Assistant ​

yaml
models:
  - name: "code-assistant"
    provider: "mistral"
    model: "codestral-latest"
    apiKey: "${MISTRAL_API_KEY}"
    roles: ["edit", "apply"]
    defaultCompletionOptions:
      temperature: 0.2
      maxTokens: 4096

Fast Response ​

yaml
models:
  - name: "fast-response"
    provider: "mistral"
    model: "mistral-small-latest"
    apiKey: "${MISTRAL_API_KEY}"
    roles: ["chat", "autocomplete"]
    defaultCompletionOptions:
      temperature: 0.5
      maxTokens: 2048

Model Features ​

Mistral Large ​

  • Parameters: 123B
  • Context Length: 32k tokens
  • Strengths: Powerful reasoning and multilingual capabilities
  • Use Cases: Complex tasks, high-quality generation

Mistral Medium ​

  • Context Length: 32k tokens
  • Strengths: Balanced performance and cost
  • Use Cases: General tasks

Mistral Small ​

  • Context Length: 32k tokens
  • Strengths: Fast response
  • Use Cases: Lightweight tasks

Codestral ​

  • Parameters: 22B
  • Context Length: 32k tokens
  • Strengths: Specialized code generation capabilities
  • Use Cases: Programming assistance

Multilingual Support ​

Mistral models support multiple languages:

  • English
  • French
  • German
  • Spanish
  • Italian
  • Portuguese
  • Chinese (limited support)

Troubleshooting ​

Common Errors ​

  1. 401 Unauthorized: Invalid API key
  2. 429 Too Many Requests: Rate limit reached
  3. 400 Bad Request: Parameter error
  4. Model Not Found: Incorrect model name

Debugging Steps ​

  1. Verify API key format and validity
  2. Check model name spelling
  3. Monitor usage and rate limits
  4. View error response details
  5. Confirm account balance is sufficient

Best Practices ​

1. Model Selection ​

  • High Quality Needs: Use Mistral Large
  • Balanced Cost: Use Mistral Medium
  • Fast Response: Use Mistral Small
  • Code Tasks: Use Codestral
  • Open-Source Priority: Use Open model series

2. Parameter Optimization ​

  • Lower temperature for more consistent results
  • Set reasonable maxTokens limits
  • Use stop sequences to control output
  • Adjust penalty parameters to avoid repetition

3. Cost Control ​

  • Monitor token usage
  • Choose appropriate model size
  • Optimize prompt length
  • Set quota alerts

4. Performance Optimization ​

  • Use streaming responses for better UX
  • Implement request caching
  • Batch similar requests
  • Choose appropriate model version

Copyright © 2025 ByteBuddy