Frontier model API

How xAI Grok API compares

unverifiedcatalog v1

Grok API. Short-context rates below 200k prompt tokens, plus the published rate limits.

Pricing page · Site · Open in compare

Regions and compliance

Regions: us-east-1, us-west-2, us-central-1

Compliance: not published / unverified

Token offerings

Grok 4.7 (prompt under 200k)

serverless_api · grok-4.7 · open-weight not published / unverified

Input / 1M
2docs.x.aichecked 2026-10-20“Input | < 200k prompt tokens | $2.00”
Cached input / 1M
0.5docs.x.aichecked 2026-10-20“Cached input | < 200k prompt tokens | $0.50”
Output / 1M
6docs.x.aichecked 2026-10-20“Output | < 200k prompt tokens | $6.00”

Context 500,000 · tokens/s not published · TTFT not published

At or above 200k prompt tokens the same page lists $4 input, $1 cached, $12 output. Batch API: not supported.

Compute SKUs

No GPU or node price is seeded.

Propose a correction

This goes to a moderation queue. It changes the catalog only after approval, and the public changelog records who supplied it.

Current value: unverified

How xAI Grok API compares · AgentTune