AI Provider Setup · BYOK

Connect xAI (Grok)

Grok 4 with real-time X/Twitter access.

xAI's Grok models have built-in access to real-time X (Twitter) data, which no other frontier lab offers. Good pick when questions depend on what's happening on X right now — brand mentions, trending topics, influencer activity. Otherwise, Grok is a solid general-purpose model competitive with GPT-5 and Claude.

Best used for

Key details

Where to generate your key
https://console.x.ai
Expected key format
Starts with xai-
Environment variable name (if self-hosting)
XAI_API_KEY
Provider pricing
https://docs.x.ai/docs/models
Official docs
xAI docs

Step-by-step setup

  1. 1 Create an xAI Console account

    Go to console.x.ai and sign up. xAI is currently invite-friendly — most sign-ups are approved immediately.

  2. 2 Add billing

    Open Billing and add a card. xAI requires an active payment method before any API key works.

  3. 3 Navigate to API Keys

    In console.x.ai, open the API Keys tab. xAI lets you create multiple keys with different ACL scopes.

  4. 4 Create a key labeled Admaxxer

    Click Create API Key, name it Admaxxer, and grant it the models and endpoints permissions. The key starts with xai- and is shown once.

  5. 5 Paste into Admaxxer

    Open Settings then AI Providers, find the xAI (Grok) row, paste the key, and click Connect.

  6. 6 Pick a default model

    grok-4 is our recommended pick — best reasoning and the flagship tier. Use grok-4-fast-reasoning for faster chain-of-thought, grok-3 if you need an older stable pin.

  7. 7 If the connection test fails

    Confirm the key starts with xai-, billing is active, and the key has the chat and models scopes granted.

Test the connection inside Admaxxer

Once you've pasted your key into /settings/ai-providers, hit Test. Admaxxer makes a single no-cost call against the provider's /v1/models (or equivalent) endpoint and reports back:

Available models

These are the pinned model IDs Admaxxer tests against. The model picker in chat will show these plus any live-catalog entries the provider returns.

Model ID Display name Context Tools Best for
grok-4 Grok 4 (recommended) 256K Yes Flagship — best reasoning, 256K context, real-time X access. Default pick.
grok-4-fast-reasoning Grok 4 Fast Reasoning 256K Yes Faster reasoning-tuned variant. Good for agent loops where latency matters.
grok-3 Grok 3 131K Yes Previous-gen flagship. Use as a stable pin if you need reproducibility across deploys.

Common errors and fixes

401

When it happens: On Connect if the key is wrong, revoked, or missing the needed scopes.

Fix: Regenerate at console.x.ai and grant it the chat and models scopes on creation.

402

When it happens: If billing is off or the card was declined.

Fix: Open console.x.ai/billing and confirm the payment method is active.

429

When it happens: Under concurrent load on grok-4.

Fix: Request a rate-limit increase in xAI console, or drop to grok-4-fast-reasoning.

model_not_found

When it happens: If you pick a retired Grok build.

Fix: Use grok-4, grok-4-fast-reasoning, or grok-3 — the current supported tiers.

timeout

When it happens: During peak demand — xAI capacity varies.

Fix: Retry with backoff. For mission-critical flows, fail over to a secondary provider via OpenRouter.