PRODUCT / INTEGRATIONS

Connect the resources your workforce actually needs.

Model providers, channels and browser capability are configured through the connected runtime. This directory describes connection types represented in the current AgentOS product—not a claim that every resource is active in every deployment.

Model providerConfigured

OpenAI

Available when configured in your AgentOS and OpenClaw environment.

Model providerConfigured

Anthropic

Available when configured in your AgentOS and OpenClaw environment.

Model providerConfigured

Gemini

Available when configured in your AgentOS and OpenClaw environment.

Model providerConfigured

xAI

Available when configured in your AgentOS and OpenClaw environment.

ChannelRuntime-dependent

Telegram

A workspace surface supported through OpenClaw where configured.

ChannelRuntime-dependent

Discord

A workspace surface supported through OpenClaw where configured.

ChannelRuntime-dependent

Slack

A workspace surface supported through OpenClaw where configured.

BrowserExperimental

Browser profiles

An MVP bridge through the underlying runtime; typed browser-profile dispatch is not yet exposed.

Connection availability depends on your configured provider keys, Gateway capabilities, scopes and account policies. AgentOS makes unsupported or degraded paths visible rather than representing them as ready.

Connect a model before connecting everything else.

A single ready provider and a scoped task are the clearest way to verify a new deployment.

Read deployment guide