ChatGPT
OpenAIAPI keyOAuth32 models
How to route to it
Any of these in the model field reaches this provider. The explicit form is the one to prefer — prefix inference is a convenience, and a convenience is the first thing to become ambiguous when a provider adds a model family.
- chatgpt/<model>
- openai/<model>
- gpt/<model>
How it authenticates
- API keyPaste an API key in the dashboard. It is stored in the local SQLite database and sent only to this provider.
- OAuthSign in with the subscription you already pay for. The token lives on your machine and refreshes itself; nothing about it is transmitted anywhere else.
Models it carries
- GPT-5.4gpt-5.4400K
- GPT-5.4 Minigpt-5.4-mini400K
- GPT-5.4 Nanogpt-5.4-nano400K
- GPT-5.2gpt-5.2400K
- GPT-5.1gpt-5.1400K
- GPT-5gpt-5400K
- GPT-5 Minigpt-5-mini400K
- GPT-5 Nanogpt-5-nano400K
- GPT-4ogpt-4o128K
- GPT-4o Minigpt-4o-mini128K
- GPT-4 Turbogpt-4-turbo128K
- GPT-4.1gpt-4.1128K
- GPT-4.1 Minigpt-4.1-mini128K
- GPT-4.1 Nanogpt-4.1-nano128K
- O3o3200K
- O3 Minio3-mini200K
- O3 Proo3-pro200K
- O4 Minio4-mini200K
- O1o1200K
- O1 Minio1-mini200K
- Text Embedding 3 Largetext-embedding-3-large200K
- Text Embedding 3 Smalltext-embedding-3-small200K
- Text Embedding Ada 002text-embedding-ada-002200K
- TTS-1tts-1200K
- TTS-1 HDtts-1-hd200K
- GPT-4o Mini TTSgpt-4o-mini-tts128K
- Whisper 1whisper-1200K
- GPT-4o Transcribegpt-4o-transcribe128K
- GPT-4o Mini Transcribegpt-4o-mini-transcribe128K
- GPT Image 1gpt-image-1200K
- DALL-E 3dall-e-3200K
- DALL-E 2dall-e-2200K