Open Models Configuration

RA.Aid supports a variety of open source and compatible model providers. This guide covers configuration options and best practices for using different models with RA.Aid.

Overview

Supported Providers

RA.Aid supports these model providers:

Provider	Description	Key Features
DeepSeek	Chinese hedge fund who creates sophisticated LLMs	Strong, open models like R1
Fireworks	Serverless AI inference platform for open-source models	High-performance inference, pay-per-token, variety of open models
MakeHub	OpenAI-compatible AI router optimized for developers	Customizable speed/price ratio, curated latest models, dev-optimized performance
OpenRouter	Multi-model gateway service	Access to 100+ models, unified API interface, pay-per-token
OpenAI-compatible	Self-hosted model endpoints	Compatible with Llama, Mistral and other open models
Anthropic	Claude model series	200k token context, strong tool use, JSON/XML parsing
Gemini	Google's multimodal models	Code generation in 20+ languages, parallel request support
Ollama	Local LLM hosting framework	Run models locally, no API keys required, offline usage
Bedrock	Amazon LLM platform on AWS	Large variety of models, serverless, leverages IAM policies

Quick Setup

Basic Configuration

1. Set your provider's API key:

# Choose the appropriate provider
export DEEPSEEK_API_KEY=your_key
export FIREWORKS_API_KEY=your_key
export MAKEHUB_API_KEY=your_key
export OPENROUTER_API_KEY=your_key
export OPENAI_API_KEY=your_key
export ANTHROPIC_API_KEY=your_key
export GEMINI_API_KEY=your_key

export AWS_PROFILE=aws_profile_name # For Bedrock
# Alternative configurations for Bedrock
export AWS_SECRET_ACCESS_KEY=your_aws_secret_key
export AWS_ACCESS_KEY_ID=your_aws_key_id

# Ollama doesn't require an API key

2. Run RA.Aid with your chosen provider:

ra-aid -m "Your task" --provider <provider> --model <model>

Provider Configuration

DeepSeek

DeepSeek Models

DeepSeek offers powerful reasoning models optimized for complex tasks.

# Environment setup
export DEEPSEEK_API_KEY=your_api_key_here

# Basic usage
ra-aid -m "Your task" --provider deepseek --model deepseek-reasoner

# With temperature control
ra-aid -m "Your task" --provider deepseek --model deepseek-reasoner --temperature 0.7

Available Models:

• deepseek-reasoner: Optimized for reasoning tasks
• Access via OpenRouter: deepseek/deepseek-r1

MakeHub

MakeHub Integration

MakeHub is an OpenAI-compatible AI router specifically optimized for developers, offering access to curated latest models with customizable speed/price ratios.

# Environment setup
export MAKEHUB_API_KEY=your_api_key_here

# Basic usage
ra-aid -m "Your task" --provider makehub --model anthropic/claude-4-sonnet
ra-aid -m "Your task" --provider makehub --model qwen/Qwen2.5-Coder-32B
ra-aid -m "Your task" --provider makehub --model deepseek/deepseek-V3-0324-fp8

# With temperature control
ra-aid -m "Your task" --provider makehub --model anthropic/claude-4-sonnet --temperature 0.3

Why MakeHub for Developers:

• Curated Latest Models: Access to hand-picked cutting-edge models optimized for development
• Customizable Speed/Price: Configure the optimal balance for your workload via price_performance_ratio
• Dev-Optimized Performance: Intelligent routing tuned specifically for development tasks
• OpenAI-Compatible: Drop-in replacement for OpenAI API calls with familiar interface

Available Models:

• anthropic/claude-4-sonnet: Latest Claude model optimized for coding tasks

• qwen/Qwen2.5-Coder-32B: Specialized for code generation and development

• deepseek/deepseek-V3-0324-fp8: High-performance reasoning model for complex problems

• mistral/mistral-small-24B-fp16: Efficient model for rapid development workflows

• google/gemini-2.5-pro-preview: Google's latest multimodal model with coding capabilities

• look at 'https://makehub.ai/models' for more models

Configuration Options:

• Supports temperature control and all standard OpenAI parameters
• Configure price_performance_ratio for intelligent speed/cost optimization
• Automatic routing to optimal endpoints - no manual base URL configuration needed
• Expert mode support with EXPERT_MAKEHUB_API_KEY environment variable

OpenRouter

OpenRouter Integration

OpenRouter provides access to multiple open source models through a single API.

# Environment setup
export OPENROUTER_API_KEY=your_api_key_here

# Example commands
ra-aid -m "Your task" --provider openrouter --model mistralai/mistral-large-2411
ra-aid -m "Your task" --provider openrouter --model deepseek/deepseek-r1

Popular Models:

• mistralai/mistral-large-2411
• anthropic/claude-3
• deepseek/deepseek-r1

OpenAI-compatible

OpenAI-compatible Endpoints

Use OpenAI-compatible API endpoints with custom hosting solutions.

# Environment setup
export OPENAI_API_KEY=your_api_key_here
export OPENAI_API_BASE=https://your-api-endpoint

# Usage
ra-aid -m "Your task" --provider openai-compatible --model your-model-name

Configuration Options:

• Set custom base URL with OPENAI_API_BASE
• Supports temperature control
• Compatible with most OpenAI-style APIs

Google Gemini

Google Gemini Models

Google's Gemini models offer powerful multimodal capabilities with extensive code generation support.

# Environment setup
export GEMINI_API_KEY=your_api_key_here

# Basic usage
ra-aid -m "Your task" --provider gemini --model gemini-1.5-pro-latest

# With temperature control
ra-aid -m "Your task" --provider gemini --model gemini-1.5-flash-latest --temperature 0.5

Available Models:

• gemini-pro: Original Gemini Pro model
• gemini-1.5-flash-latest: Latest Gemini 1.5 Flash model (fast responses)
• gemini-1.5-pro-latest: Latest Gemini 1.5 Pro model (strong reasoning)
• gemini-1.5-flash: Gemini 1.5 Flash release
• gemini-1.5-pro: Gemini 1.5 Pro release
• gemini-1.0-pro: Original Gemini 1.0 Pro model

Configuration Notes:

• All Gemini models support a 128,000 token context window
• Temperature control is supported for creative vs. deterministic responses
• Obtain your API key from AI Studio

Ollama

Ollama Integration

Ollama provides a framework for running large language models locally on your machine.

# No environment variables required by default
# Ollama uses http://localhost:11434 by default

# Basic usage
ra-aid -m "Your task" --provider ollama --model justinledwards/mistral-small-3.1-Q6_K

# With context window control
ra-aid -m "Your task" --provider ollama --model qwq:32b --num-ctx 8192

# With temperature control
ra-aid -m "Your task" --provider ollama --model MHKetbi/Qwen2.5-Coder-32B-Instruct --temperature 0.3

Popular Models:

• justinledwards/mistral-small-3.1-Q6_K: Optimized Mistral small model
• qwq:32b: High-performing yet compact reasoning model
• MHKetbi/Qwen2.5-Coder-32B-Instruct: Qwen 2.5 optimized for code tasks

Configuration Notes:

• Requires Ollama to be installed and running
• No API keys needed - all processing happens locally
• Adjust context window with --num-ctx (default: 262144)
• Works completely offline after model download

For detailed setup instructions and advanced configuration options, see our Ollama Configuration Guide.

Bedrock

Bedrock Integration

Bedrock provides access to a variety of models hosted on AWS.

# Basic usage
ra-aid -m "Your task" --provider bedrock --model us.anthropic.claude-3-7-sonnet-20250219-v1:0

# Expert usage
ra-aid -m "Your task" --provider bedrock --model claude-3-5-haiku-latest --expert-provider bedrock --expert-model us.anthropic.claude-3-7-sonnet-20250219-v1:0

Configuration Notes:

• Requires AWS credentials configured in your environment
• Set AWS_PROFILE for named profiles or use AWS_ACCESS_KEY_ID, AWS_SECRET_ACCESS_KEY, and AWS_SESSION_TOKEN for direct access
• Use AWS_REGION to specify the region (default: us-east-1)
• Some models require cross-region inference identifier instead of the model name (e.g., us.anthropic.claude-3-7-sonnet-20250219-v1:0)

Advanced Configuration

Expert Model

Expert Tool Configuration

Configure the expert model for specialized tasks; this usually benefits from a more powerful, slower, reasoning model:

# DeepSeek expert
export EXPERT_DEEPSEEK_API_KEY=your_key
ra-aid -m "Your task" --expert-provider deepseek --expert-model deepseek-reasoner

# MakeHub expert
export EXPERT_MAKEHUB_API_KEY=your_key
ra-aid -m "Your task" --expert-provider makehub --expert-model gpt-4o

# OpenRouter expert
export EXPERT_OPENROUTER_API_KEY=your_key
ra-aid -m "Your task" --expert-provider openrouter --expert-model mistralai/mistral-large-2411

# Gemini expert
export EXPERT_GEMINI_API_KEY=your_key
ra-aid -m "Your task" --expert-provider gemini --expert-model gemini-2.0-flash-thinking-exp-1219

# Ollama expert
ra-aid -m "Your task" --expert-provider ollama --expert-model qwq:32b

Temperature Control

Temperature Settings

Control model creativity vs determinism:

# More deterministic (good for coding)
ra-aid -m "Your task" --temperature 0.2

# More creative (good for brainstorming)
ra-aid -m "Your task" --temperature 0.8

Note: Not all models support temperature control. Check provider documentation.

Best Practices

• Set environment variables in your shell configuration file
• Use lower temperatures (0.1-0.3) for coding tasks
• Test different models to find the best fit for your use case
• Consider using expert mode for complex programming tasks

Environment Variables

Complete list of supported environment variables:

Variable	Provider	Purpose
OPENROUTER_API_KEY	OpenRouter	Main API access
DEEPSEEK_API_KEY	DeepSeek	Main API access
FIREWORKS_API_KEY	Fireworks	Main API access
MAKEHUB_API_KEY	MakeHub	Main API access
OPENAI_API_KEY	OpenAI-compatible	API access
OPENAI_API_BASE	OpenAI-compatible	Custom endpoint
ANTHROPIC_API_KEY	Anthropic	API access
GEMINI_API_KEY	Gemini	API access
AWS_PROFILE	Bedrock	API access
AWS_SECRET_ACCESS_KEY	Bedrock	API access
AWS_ACCESS_KEY_ID	Bedrock	API access
AWS_SESSION_TOKEN	Bedrock	API access
AWS_REGION	Bedrock	Region specification (default: us-east-1)
OLLAMA_BASE_URL	Ollama	Custom endpoint (default: http://localhost:11434)
EXPERT_OPENROUTER_API_KEY	OpenRouter	Expert tool
EXPERT_DEEPSEEK_API_KEY	DeepSeek	Expert tool
EXPERT_FIREWORKS_API_KEY	Fireworks	Expert tool
EXPERT_MAKEHUB_API_KEY	MakeHub	Expert tool
EXPERT_GEMINI_API_KEY	Gemini	Expert tool
EXPERT_OLLAMA_BASE_URL	Ollama	Expert tool endpoint
EXPERT_AWS_PROFILE	Bedrock	Expert tool
EXPERT_AWS_SECRET_ACCESS_KEY	Bedrock	Expert tool
EXPERT_ACCESS_KEY_ID	Bedrock	Expert tool
EXPERT_AWS_SESSION_TOKEN	Bedrock	Expert tool

Troubleshooting

• Verify API keys are set correctly
• Check endpoint URLs for OpenAI-compatible setups
• Monitor API rate limits and quotas
• For Ollama, ensure the service is running (ollama list)

See Also

• Expert Model Configuration - Detailed information about expert model setup and configuration
• Thinking Models - Learn about models that can show their reasoning process