vrr/agent-zero
2
0
Fork 0
mirror of https://github.com/agent0ai/agent-zero.git synced 2026-05-17 04:01:13 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 2
agent-zero/plugins/_model_config
History
Alessandro 7c59ac9e57 Show active models for Default LLM 
Expose sanitized active main and utility model metadata through the model override endpoint, then render those names in the chat model switcher even when no preset override is active. Keep the inline model names hidden on narrow screens and cover the behavior with a regression check.

Refresh model names after settings save

Refresh the active chat model switcher after _model_config settings are saved so changes to main and utility models appear immediately. Extend the model switcher regression check to cover the save-refresh hook.
2026-04-28 16:14:53 +02:00
..
api Show active models for Default LLM 2026-04-28 16:14:53 +02:00
extensions Show active models for Default LLM 2026-04-28 16:14:53 +02:00
helpers Add project-scoped LLM presets 2026-04-21 18:18:59 +02:00
webui Show active models for Default LLM 2026-04-28 16:14:53 +02:00
default_config.yaml Update default_config.yaml 2026-03-24 19:28:32 +01:00
default_presets.yaml Refactor default presets, fix chardet version conflict, improve welcome banner refresh logic 2026-03-24 19:17:12 +01:00
hooks.py Refactor API key handling and update system improvements 2026-03-24 14:41:46 +01:00
plugin.yaml feat: add _model_config plugin with call-time model resolution 2026-03-14 09:41:19 -07:00
README.md browser: replace browser-use agent with native browser 2026-04-24 15:43:52 +02:00

README.md

Model Configuration

Manage which models Agent Zero uses for chat, utility, and embeddings, with support for scoped overrides and reusable presets.

What It Does

This plugin centralizes model selection and model-related settings for the application. It provides helpers and APIs for:

  • selecting chat, utility, and embedding models
  • reading and saving model presets
  • checking for missing API keys
  • allowing optional per-chat model overrides
  • resolving config at global, project, agent, and chat scope

Main Behavior

  • Scoped configuration
    • Reads plugin config through the standard plugin config system with project and agent overrides.
  • Preset management
    • Loads presets from a user file when present and falls back to bundled defaults.
    • Project presets can be stored beside a project's scoped model config.
  • Per-chat override
    • Allows a chat context to store a temporary override or preset reference in context data.
  • Model object construction
    • Builds ModelConfig objects and the runtime chat, utility, and embedding wrappers used elsewhere in the app.
  • API key validation
    • Reports configured providers that still require API keys.

Key Files

  • Core helper
    • helpers/model_config.py resolves config, presets, overrides, and runtime model objects.
  • APIs
    • api/model_config_get.py
    • api/model_config_set.py
    • api/model_override.py
    • api/model_presets.py
    • api/model_search.py
    • api/api_keys.py
  • Hooks
    • hooks.py exposes plugin-level integration hooks.

Configuration Scope

  • Settings section: agent
  • Per-project config: true
  • Per-agent config: true
  • Always enabled: true

Project-Scoped Model Config

Projects store copied model settings in the standard scoped plugin path:

/a0/usr/projects/<project>/.a0proj/plugins/_model_config/config.json

Project-only presets live beside that config:

/a0/usr/projects/<project>/.a0proj/plugins/_model_config/presets.yaml

The project preset file uses the same plain YAML list schema as global presets. It does not contain scope metadata:

- name: Research
  chat:
    provider: openrouter
    name: anthropic/claude-sonnet-4.6
    api_base: ""
    ctx_length: 200000
    ctx_history: 0.7
    vision: true
  utility:
    provider: openrouter
    name: openai/gpt-5.4-mini
    api_base: ""
    ctx_length: 128000
    ctx_input: 0.7

Selecting a preset for a project copies the preset's chat and optional utility settings into the project's config.json. The embedding model is copied from the current effective config, because presets currently define chat and utility only.

Plugin Metadata

  • Name: _model_config
  • Title: Model Configuration
  • Description: Manages LLM model selection and configuration for chat, utility, and embedding models. Supports per-project and per-agent overrides with optional per-chat model switching.