Model Catalogue
ChatBotKit gives you access to a broad catalogue of AI models from leading providers, all behind one consistent bot configuration. New models are added as they are released, so you can run the latest from providers like OpenAI, Anthropic, Google, Mistral, DeepSeek, Groq, and Perplexity without managing separate integrations for each.
Every model is assigned to a bot the same way, which means you can choose the best fit for each use case - a high-reasoning model for complex agents, a fast model for high-volume chat, or a realtime model for voice - and switch any time.
What You Can Do
- Choose per bot: Assign any supported model to a bot from a single configuration screen.
- Use many providers: Run models from leading AI providers without separate setup for each.
- Stay current: New models are added as they become available across providers.
- Bring your own: Pair the catalogue with Bring Your Own Key and custom models to run fine-tuned or self-licensed models.
- Match the task: Select for reasoning depth, speed, cost, context length, or realtime voice.
How It Works
Models share the same bot configuration workflow, so changing a bot's model is a single selection - the rest of your setup, including backstory, skillsets, datasets, and memory, stays the same. This keeps experimentation cheap: try a new model on a bot, compare behavior, and switch back if needed, all without rebuilding anything.
Setup
Open a bot's configuration, select a model from the catalogue, and save. To use a self-licensed or fine-tuned model, add it through Bring Your Own Key and it appears alongside the built-in options.
Practical Uses
A broad model catalogue lets you optimize each bot independently: route a customer-facing assistant to a fast, cost-efficient model, give a research agent a high-reasoning model, and run voice experiences on a realtime model - all in the same account, with the freedom to adopt new models as they ship.