Skip Grok 4.1 for Agentic Work
Date: 2026-02-22 Version: OpenClaw 2026.2.21-2 (35a57bc)
xai/grok-4-1-fast-reasoning is not reliable for agentic tasks. Common failure modes: claiming to have read a file without doing so, reporting completed actions that never happened, and fabricating errors. The hallucinations are frequent enough to make it unusable for any multi-step work where you need to trust tool results.
Stick to Claude (Sonnet/Opus) or GPT Codex 5.2+ for anything agentic.
OpenClaw Config
"xai": {
"baseUrl": "https://api.x.ai/v1",
"apiKey": "X",
"api": "openai-responses",
"models": [
{
"id": "grok-4-1-fast-reasoning",
"name": "Grok 4.1 Fast Reasoning",
"api": "openai-responses",
"reasoning": false,
"input": ["text"],
"cost": { "input": 0, "output": 0, "cacheRead": 0, "cacheWrite": 0 },
"contextWindow": 2000000,
"maxTokens": 8192
}
]
}