Supported Providers & Models

🚀 Janito is optimized and tested for the default model: gpt-4.1. 🧪 Testing and feedback for other models is welcome!

🤖 Model Types

Janito is compatible with most OpenAI-compatible chat models, including but not limited to:

• gpt-4.1 (default)
• Azure-hosted OpenAI models (with correct deployment name)
• Google Gemini models (e.g., gemini-2.5-flash)

🛠️ How to Select a Model

• Use the --model CLI option to specify the model for a session: janito "Prompt here" --model gpt-4.1
• Configure your API key and endpoint in the configuration file or via CLI options.

📋 Supported Models Table

Model	Status	Context	Max Input	Max CoT	Max Response	Thinking	Provider	Reference
gpt-3.5-turbo	Supported	16385	12289	N/A	4096		OpenAI	source
gpt-4.1	Supported	1047576	1014808	N/A	32768		OpenAI	source
gpt-4.1-mini	Supported	1047576	1014808	N/A	32768		OpenAI	source
gpt-4.1-nano	Supported	1047576	1014808	N/A	32768		OpenAI	source
gpt-4-turbo	Supported	128000	N/A	N/A	N/A		OpenAI	source
gpt-4o	Supported	128000	123904	N/A	4096		OpenAI	source
gpt-4o-mini	Supported	128000	111616	N/A	16384		OpenAI	source
o3-mini	Supported	200000	100000	N/A	100000	📖	OpenAI	source
o3	Supported	200000	100000	N/A	100000	📖	OpenAI	source
o4-mini	Supported	200000	100000	N/A	100000	📖	OpenAI	source
gemini-2.5-flash	Supported	N/A	N/A	24576	8192	✔️	Google	source

Context window: 200 k tokens
Max input: 100 k tokens
Max CoT: N/A
Max response: 100 k tokens
Thinking: 📖
Driver: OpenAI

ℹ️ Notes

• Some advanced features (like tool calling) require models that support OpenAI function calling.
• Model availability and pricing depend on your provider and API key.
• For the latest list of supported models, see your provider’s documentation or the OpenAI models page and Google Gemini documentation.