xAI has announced the retirement of several earlier models from its API, effective May 15, 2026, at 12:00 PM PT. This move is intended to focus development on the newest generation of Grok models. After this date, requests to the retired model slugs will automatically redirect to Grok 4.3 or Grok Build 0.1, depending on the original model.
Key Points
- xAI is retiring several models from its API on May 15, 2026, at 12:00 PM PT.
- Retired model slugs will automatically redirect to Grok 4.3 or Grok Build 0.1 after the retirement date.
- Grok 4.3 is priced at $1.25 per 1M input tokens and $2.50 per 1M output tokens.
- Users continuing to send requests to deprecated slugs after May 15, 2026, will be billed at Grok 4.3 pricing.
- Grok 4.3 offers four reasoning effort levels: none, low, medium, and high.
- Models like grok-4-1-fast-reasoning, grok-4-fast-reasoning, and grok-4-0709 are recommended to migrate to Grok 4.3.
- grok-code-fast-1 is recommended to migrate to grok-build-0.1 for improved coding capabilities.
Context
According to xAI, the retirement of earlier models allows the company to concentrate on advancing its newest generation of Grok models. The slugs for the retired models will continue to resolve, meaning existing code will not break, but pricing and model behavior will change.
Why It Matters
Developers using xAI's API need to be aware of the upcoming model retirements and the automatic redirection to avoid unexpected cost increases or changes in model performance. Explicitly migrating to the recommended models allows for better control over pricing and reasoning effort.
What To Do
- Note the May 15, 2026, retirement date for several xAI models.
- Review your API requests to identify any use of the models being retired.
- Explicitly update your code to specify Grok 4.3 or Grok Build 0.1 before the retirement date.
- Compare the pricing of Grok 4.3 with your current model's rates to anticipate cost changes.
- Test workloads with Grok 4.3 using different reasoning effort levels (none, low, medium, high) to optimize performance and cost.
