Skip to main content
The behaviors library documentation is still under construction.
behaviors is a Python library for building AI evaluations. It provides model adapters, multi-turn conversations, user and tool policies, and structured conversation records. Use it to sample a model, simulate an interaction, or ask a model to judge another model’s output. You control the messages, tools, and stopping conditions of an interaction. behaviors handles the conversation loop and provides configurable retries and failure recording. Users can be scripted, or simulated by models: a base model writing the user’s messages, optionally steered by a chat model (see Simulated users). behaviors supplies no general-purpose judge: you write one for your experiment from its interfaces, as the judge step does. behaviors works with fxtr: conversation records can be stored as fxtr entities, returned from steps, and inspected in the experiment viewer. fxtr manages the experiment’s execution, caching, and provenance. For the shortest example of a step that calls a model, see Steps in the fxtr concepts.

Calling language models

Set up behaviors, run conversations, and judge their outputs in fxtr steps.

Install fxtr

Set up the project that runs and stores your experiments.
The other guides go further: choosing models, policies and tools, rollout outcomes and retries, conversation records and checkpoints, and Docent readings. The API overview maps the library’s layers and types.