Overview
Talemate uses client(s) to connect to local or remote AI text generation APIs like koboldcpp, text-generation-webui or OpenAI.
The various agents will use these connections to generate content.
Why multiple clients?
Talemate supports and encourages use of multiple clients. We believe it makes sense to give certain tasks to certain models. Some models may be more suited for for management tasks, while others may be better at creative tasks.
It is, however, perfectly fine to just use a single client for all tasks if you prefer.
What Talemate expects from a model
Talemate expects a current reasoning-capable model with reliable JSON and tool-call output and a usable context of 16k tokens or more. The Model expectations page states the bar, lists known good examples for hosted and self-hosted setups, and describes the symptoms of a model below it.
Checking a model
The bundled Model Testing Harness scene runs a fixed suite of minimum-viability tests against the clients assigned to your agents. Load it after configuring a new client to confirm the model can handle what Talemate will ask of it.
Client setup instructions
Officially supported APIs
Remote APIs
Local APIs
APIs functional via OpenAI compatible client
Note
These APIs do not have a talemate client, but the OpenAI compatible client can be used to connect to them.