TokenCost lists Grok 4.1 Fast at $0.20 per million input tokens and $0.50 per million output tokens, matching xAI’s launch pricing, but xAI’s own documentation says the Grok 4.1 Fast API variants were retired on May 15, 2026 and now redirect to Grok 4.3 billing.
xAI’s documentation says the Grok 4.1 Fast API variants were retired on May 15, 2026, creating an important caveat for developers comparing current pricing pages for the model.
TokenCost’s page for “Grok 4.1 Fast API Pricing,” marked as updated in June 2026 and verified on June 14, lists Grok 4.1 Fast at $0.200 per 1 million input tokens and $0.500 per 1 million output tokens. The same page lists a 2.0 million-token context window.
Those figures match the pricing described in xAI’s launch post for Grok 4.1 Fast. In that announcement, xAI said Grok 4.1 Fast was available through the API in both reasoning and non-reasoning variants, with a 2 million-token context window, priced at $0.20 per 1 million input tokens and $0.50 per 1 million output tokens.
On those two sources alone, the historical launch pricing appears consistent: both TokenCost and xAI’s launch post give the same input and output token rates and the same context-window size.
The more important issue is current availability. In xAI Docs’ “May 15, 2026 Model Retirement” notice, xAI says grok-4-1-fast-reasoning and grok-4-1-fast-non-reasoning were retired from the API effective May 15, 2026. The same xAI documentation says those retired model names now redirect to grok-4.3 billing.
That means developers should not treat the older Grok 4.1 Fast launch price as necessarily applying to current API usage if they are calling the retired model identifiers. According to xAI’s own documentation, those identifiers no longer represent independent Grok 4.1 Fast billing after the May 15 retirement date.
For teams estimating API costs, the distinction is material. TokenCost’s listed price may still be useful as a record of the Grok 4.1 Fast launch price, and it is consistent with xAI’s original announcement. But xAI Docs indicates that current billing behavior changed after the model retirement.
Before using the TokenCost figures in a budget, developers should check the active xAI model documentation and billing details for grok-4.3, especially if their application still references grok-4-1-fast-reasoning or grok-4-1-fast-non-reasoning. The retirement notice indicates those calls are now redirected, so the relevant price for new cost estimates may be the redirected model’s billing rather than the retired model’s launch rate.
TokenCost and xAI’s launch post agree that Grok 4.1 Fast launched with a 2 million-token context window and pricing of $0.20 per million input tokens and $0.50 per million output tokens. However, xAI Docs says the Grok 4.1 Fast reasoning and non-reasoning API variants were retired on May 15, 2026 and now redirect to Grok 4.3 billing. For present-day API planning, xAI’s retirement documentation is the key source to consult.
xAI’s documentation says the Grok 4.1 Fast API variants were retired on May 15, 2026, creating an important caveat for developers comparing current pricing pages for the model.
Those figures match the pricing described in xAI’s launch post for Grok 4.1 Fast.
On those two sources alone, the historical launch pricing appears consistent: both TokenCost and xAI’s launch post give the same input and output token rates and the same context window size.
Continue reading
