​

Supported Models

Index supports several powerful large language models with vision capabilities:

​

Gemini

​

Gemini 2.5 Pro

• Status: High performance and fast execution
• Usage: Best choice for complex reasoning tasks that require fast execution
• How to use:

  from index import Agent, GeminiProvider
  
  llm = GeminiProvider(
          model="gemini-2.5-pro-preview-03-25",
          thinking_token_budget=8192
        )
  

​

Gemini 2.5 Flash

• Status: Good performance and fast execution, very cost effective
• Usage: Best choice for simpler tasks that require fast execution and low cost
• How to use:

  from index import Agent, GeminiProvider
  
  llm = GeminiProvider(
          model="gemini-2.5-flash-preview-04-17",
          thinking_token_budget=8192
        )
  

​

Anthropic

​

Claude 3.7 Sonnet

• Status: High performance
• Usage: Best choice for complex reasoning tasks
• How to use:

  from index import Agent, AnthropicProvider
  
  llm = AnthropicProvider(
      model="claude-3-7-sonnet-20250219",
      enable_thinking=True,
      thinking_token_budget=2048
  )
  

​

OpenAI

​

o4-mini

• Status: Good performance
• Usage: Various difficulty tasks, depends on reasoning_effort parameter
• How to use:

  from index import OpenAIProvider
  
  llm = OpenAIProvider(model="o4-mini", reasoning_effort="low")