Skip to content

OpenAI Provider

Use OpenAI's GPT models or any OpenAI-compatible API.

Quick Start

export OPENAI_API_KEY="sk-..."
dial

Configuration

Environment Variables

export OPENAI_API_KEY="sk-..."
export OPENAI_BASE_URL="https://api.openai.com/v1"  # Optional
export OPENAI_MODEL="gpt-4"  # Optional

.env File

Create .env in your project root:

OPENAI_API_KEY=sk-...
OPENAI_BASE_URL=https://api.openai.com/v1
OPENAI_MODEL=gpt-4

Settings File

{
  "authType": "openai-compatible"
}

Available Models

Model Context Best For
gpt-4 8K Complex reasoning
gpt-4-turbo 128K Long context
gpt-4o 128K Balanced performance
gpt-3.5-turbo 16K Fast, economical

Set your model:

export OPENAI_MODEL="gpt-4o"

OpenAI-Compatible APIs

Any API that follows the OpenAI format works:

Anthropic (Claude)

export OPENAI_API_KEY="sk-ant-..."
export OPENAI_BASE_URL="https://api.anthropic.com/v1"
export OPENAI_MODEL="claude-3-opus-20240229"

Azure OpenAI

export OPENAI_API_KEY="your-azure-key"
export OPENAI_BASE_URL="https://your-resource.openai.azure.com/openai/deployments/your-deployment"
export OPENAI_MODEL="gpt-4"

OpenRouter

export OPENAI_API_KEY="your-openrouter-key"
export OPENAI_BASE_URL="https://openrouter.ai/api/v1"
export OPENAI_MODEL="openai/gpt-4"

Free models available:

export OPENAI_MODEL="qwen/qwen3-coder:free"

Together AI

export OPENAI_API_KEY="your-together-key"
export OPENAI_BASE_URL="https://api.together.xyz/v1"
export OPENAI_MODEL="meta-llama/Llama-3-70b-chat-hf"

Vision Support

For vision-capable models (GPT-4V, GPT-4o):

> [attach image] Describe this UI

Pricing

OpenAI uses pay-as-you-go pricing. See openai.com/pricing.

Track usage:

/stats

Troubleshooting

"Invalid API Key"

  • Check key format (starts with sk-)
  • Verify key is active at platform.openai.com
  • No extra whitespace

"Model Not Found"

  • Check model name spelling
  • Some models require specific access
  • Use /model to see available options

"Rate Limited"

  • OpenAI has tier-based rate limits
  • Wait and retry
  • Consider upgrading tier

"Context Length Exceeded"

/compress

Or reduce sessionTokenLimit:

{
  "sessionTokenLimit": 16000
}

Next Steps