grok-4.20-beta-0309-non-reasoning

Yes. Grok 4.20 (Non-Reasoning) applies a higher long-context tier above 199,999 input tokens. Requests at or below that boundary use the standard rate. Measure your request lengths and note which tier most of your traffic falls into before sizing costs.

Prompt caching on Grok 4.20 (Non-Reasoning) bills cached input at $0.20 per 1,000,000 token instead of $1.25 per 1,000,000 token for standard input. The saving applies to the cached prefix only; the rest of the request is billed normally.

Before switching production traffic to Grok 4.20 (Non-Reasoning), run a pilot on a representative slice of your traffic and check output quality, latency at peak concurrency, and cost per request. Verify the error and retry behaviour, and confirm the request limits you depend on match your plan.

The current model record does not publish a numeric rate limit for Grok 4.20 (Non-Reasoning). Confirm the quota for your plan in the console, then run a short load test at your expected concurrency to establish the effective ceiling before production.

Grok 4.20 (Non-Reasoning) is billed per use and the record publishes no free-tier allowance. The current published rates are input $2.0000 per 1M tokens and output $6.0000 per 1M tokens. Trial credit or promotional quota is tied to your account rather than to the model, so check the console before planning around a free tier.