Skip to content

Models & Providers ​

Providers ​

Meowbert calls models through the OpenAI Responses API. A provider is just a base URL and an API key, for example:

ProviderBase URL
OpenAIhttps://api.openai.com/v1
An API gateway or proxyIts OpenAI-compatible base URL, if it supports /v1/responses
A local model serverFor example http://192.168.1.20:8000/v1, if it supports the Responses API. Use an address the containers can reach, not localhost

The setup wizard adds your first provider. Later, manage providers in Admin Panel → AI providers. You can save several and switch between them. The selected provider is used for all platform model calls.

Chat Completions-only servers

Providers that only implement /v1/chat/completions don't work yet. Check that your provider or local server supports /v1/responses.

Models people can choose ​

Each model in the task composer's picker is an agent preset: a name, a model ID, and settings such as reasoning effort. The wizard creates one preset per model you pick, and the first becomes default, which new tasks use.

To change presets later, open Admin Panel → Model → Agent presets. It's a JSON editor. Each preset looks like this:

json
{
  "id": "default",
  "name": "GPT-5.5",
  "description": "Balanced defaults for most tasks.",
  "requiresSuperAdmin": false,
  "payload": {
    "model": "gpt-5.5",
    "responses": { "reasoning": { "effort": "medium", "summary": "auto" } }
  }
}

Leave out responses.reasoning for models that don't support reasoning. Presets can also define an Agent Swarm, where a leader model coordinates several worker models. Admin Panel → Model → Agent swarm generator builds these for you.

Background models ​

Admin Panel → Model → Specialized runtime sets the models Meowbert uses behind the scenes:

SettingUsed forIf unset
internalModelRouting connector messages to the right project, and as the fallback task modelopenai.defaultModel from the config file
fastModelTask titles, summarizing long conversations, and other quick jobs inside a taskThe task's own model
reviewerAgentThe reviewer in Quality control tasks (a preset ID)Built-in default
memorySynthesisAgentMemory refresh tasks (a preset ID)Built-in default
subagentFastAgentFast subagents (a preset ID)Not available

A small, fast model is a good choice for internalModel and fastModel.

Context windows and model details ​

Meowbert needs to know each model's context window to decide when to compact long conversations. Add entries in Admin Panel → Model → Model metadata. See Context Compaction.

Letting users bring their own provider ​

Users can connect their own API key or, experimentally, a ChatGPT subscription in their settings. Their tasks then use their provider instead of yours. See Bring Your Own Provider.