xAI (Grok)

Configuring xAI Grok models in Aimogen Pro, including the available models, vision and function calling support, and the xAI Responses API toggle.

xAI provides the Grok model family. Aimogen Pro supports Grok for text, vision, tool calling and agents.

Get a key#

1. Sign in at console.x.ai.
2. Create an API key.
3. Add credit. xAI is prepaid.

Configure#

Aimogen Pro › Settings › API Keys › xAI API Keys

One key per line, then Test API key, then save. Grok models appear in every model selector.

Models#

All current Grok models are in the plugin vision and function-calling capability lists, so they work with Chatbot Extensions, AI Agents and image input.

The xAI Responses API#

Settings › Advanced AI Settings › Enable Usage Of xAI's Responses API switches to the newer endpoint.

Leave it off unless you need something specific from it. The classic path is the more heavily exercised one in the plugin, and beta Responses implementations across providers have historically lacked streaming and tool-call support.

Reasoning models#

grok-4.20-0309-reasoning and the multi-agent variant spend extra tokens thinking before answering.

Consequences:

• Slower. A reasoning model can take several times as long as a non-reasoning one.
• More expensive. Reasoning tokens are billed.
• Better on hard tasks. Analysis, multi-step planning, structured extraction.

For bulk article writing, the non-reasoning variant is usually the better trade. For agent planning, reasoning models earn their cost.

If a reasoning model times out, raise PHP max_execution_time — see Recommended server configuration.

Common problems#

401 on a fresh key Copy error, or the key was created in a different xAI organisation.

HTTP 429 xAI rate limits by tier. Raise the request delay or add keys.

Timeouts on reasoning models Expected on shared hosting. Use a non-reasoning variant, or raise PHP limits.

Endpoint used#

https://api.x.ai/v1

Related#

• Selecting models
• AI Agents