Rawi (راوي) Documentation | Providers

Supported AI Providers in Rawi CLI

Rawi CLI supports multiple AI providers, enabling you to access a wide range of models and capabilities directly from your terminal. With minimal configuration, you can switch between providers, manage profiles, and leverage the best model for your use case.

Overview

Rawi CLI enables seamless integration with leading AI providers, including OpenAI, Anthropic, Google, Ollama (local), Azure OpenAI, Amazon Bedrock, Qwen, and xAI. Each provider offers unique models, features, and configuration options.

Provider	Models	API Key Required	Local Support
OpenAI	GPT-4o, GPT-4, GPT-3.5, O1, O3	✅	❌
Anthropic	Claude 3.5 Sonnet, Claude 4, Haiku	✅	❌
Google	Gemini 2.0 Flash, Gemini 1.5 Pro/Flash	✅	❌
Ollama	Llama 3.2, Mistral, CodeLlama, Qwen + 100+ more	❌	✅
Azure OpenAI	Enterprise OpenAI models	✅	❌
Amazon Bedrock	Claude, Llama, Titan	✅	❌
Qwen	Qwen-Max, Qwen-Plus, Qwen-Turbo	✅	❌
xAI	Grok-Beta, Grok-2	✅	❌

Key Features

• Multi-provider support: Easily switch between providers and models.
• Profile management: Store multiple configurations for different projects or use cases.
• Local and cloud models: Use local models for privacy or cloud models for advanced capabilities.
• Unified CLI experience: Consistent commands and options across all providers.

Example: Listing Supported Providers

To view all supported providers and their available models:

rawi info --providers

Sample output:

🤖 Supported AI Providers (8)

🟣 Anthropic (Claude) (anthropic)
  Models: 12 available
    • claude-4-opus-20250514
    • claude-4-sonnet-20250514
    • claude-3-7-sonnet-20250219
    ... and 9 more

🔷 Azure OpenAI (azure)
  Models: 0 available

🟠 Amazon Bedrock (bedrock)
  Models: 32 available
    • amazon.titan-tg1-large
    • amazon.titan-text-express-v1
    • amazon.titan-text-lite-v1
    ... and 29 more

🔴 Google (Gemini) (google)
  Models: 24 available
    • gemini-1.5-flash
    • gemini-1.5-flash-latest
    • gemini-1.5-flash-001
    ... and 21 more

🟢 Ollama (ollama)
  Models: 180 available
    • athene-v2
    • athene-v2:72b
    • aya-expanse
    ... and 177 more

🔵 OpenAI (GPT) (openai)
  Models: 44 available
    • o1
    • o1-2024-12-17
    • o1-mini
    ... and 41 more

🟡 Qwen (Alibaba Cloud) (qwen)
  Models: 21 available
    • qwen2.5-14b-instruct-1m
    • qwen2.5-72b-instruct
    • qwen2.5-32b-instruct
    ... and 18 more

🤖 xAI (Grok) (xai)
  Models: 19 available
    • grok-3
    • grok-3-latest
    • grok-3-fast
    ... and 16 more

Run "rawi configure --list-models <provider>" for all models

Example: Configuring a Provider

OpenAI

1. Configure OpenAI provider:

   rawi configure --provider openai --model gpt-4o --api-key sk-your-key

Ollama (local)

1. Configure Ollama (local) provider:

   rawi configure --provider ollama --model llama3.2

Example: Switching Providers

OpenAI

1. Switch to OpenAI profile:

   rawi ask "Summarize this code" --profile openai

Ollama (local)

1. Switch to Ollama profile:

   rawi ask "Summarize this code" --profile ollama

Best Practices

• Use provider-specific profiles for different projects.
• Choose local models (Ollama) for privacy-sensitive tasks.
• Use cloud providers for the latest and most capable models.
• Check available models with rawi configure --list-models <provider>.

Troubleshooting

If you encounter issues with a provider, run:

rawi info --providers
rawi configure --show --profile <profile>