OpenAI

ra ships two OpenAI providers — one for the Responses API (default) and one for the Chat Completions API.

Provider value	API	When to use
openai	Responses API	Default. Use with OpenAI directly. Supports native file inputs and structured streaming events.
openai-completions	Chat Completions API	Use with OpenAI-compatible services (Together AI, Fireworks, Groq, etc.) or if you specifically need the Chat Completions endpoint.

Setup

export OPENAI_API_KEY=sk-...
ra --provider openai "Hello"

Environment variables

Variable	Required	Description
OPENAI_API_KEY	Yes	OpenAI API key (used by both providers)

Models

Model	Notes
gpt-4.1	Flagship model
gpt-4.1-mini	Faster, cheaper
o3	Reasoning model
o4-mini	Reasoning, fast

ra --provider openai --model gpt-4.1 "Explain this error"
ra --provider openai --model o3 "Solve this step by step"

Extended thinking

Supported modes: off, low, medium, high, adaptive. Works with both providers.

ra --provider openai --thinking high "Solve this step by step"

OpenAI-compatible APIs (Chat Completions)

Most OpenAI-compatible services (Together AI, Fireworks, Groq, etc.) implement the Chat Completions endpoint, not the Responses API. Use openai-completions for these:

# Together AI
export OPENAI_API_KEY=your-together-key
ra --provider openai-completions \
  --openai-base-url https://api.together.xyz/v1 \
  --model meta-llama/Llama-3-70b-chat-hf "Hello"

# Groq
export OPENAI_API_KEY=your-groq-key
ra --provider openai-completions \
  --openai-base-url https://api.groq.com/openai/v1 \
  --model llama-3.3-70b-versatile "Hello"

# Fireworks
export OPENAI_API_KEY=your-fireworks-key
ra --provider openai-completions \
  --openai-base-url https://api.fireworks.ai/inference/v1 \
  --model accounts/fireworks/models/llama-v3p1-70b-instruct "Hello"

Or in a config file:

app:
  providers:
    openai-completions:
      baseURL: https://api.together.xyz/v1
      apiKey: ${OPENAI_API_KEY}

agent:
  provider: openai-completions
  model: meta-llama/Llama-3-70b-chat-hf

TIP

If you use --provider openai with a third-party base URL and get errors, switch to openai-completions — the service likely doesn't support the Responses API.

Custom base URL (Responses API)

If you're using a proxy or gateway that supports the OpenAI Responses API:

ra --provider openai --openai-base-url https://my-proxy.example.com/v1 "Hello"

Choosing between the two providers

Use openai (Responses API) when:

• Calling OpenAI directly
• Using a proxy that forwards to OpenAI's Responses endpoint
• You need native file attachment support

Use openai-completions (Chat Completions API) when:

• Using a third-party OpenAI-compatible service (Together, Groq, Fireworks, etc.)
• Calling an OpenAI-compatible local server (vLLM, llama.cpp server, etc.)
• The endpoint only supports /v1/chat/completions

See also

• Context Control — extended thinking details
• Configuration — provider credentials reference