Grok 4.3 non-reasoning launch and xAI API model retirement on May 15, 2026

Grok 4.3 Non-Reasoning Launch and xAI API Model Retirement Dear xAI API customers, We're excited to launch Grok 4.3, our latest flagship that leads the industry in agentic tool calling, instruction following, and non-hallucination rate. With a 1M token context window, it's our first model with configurable reasoning effort, including a new non-reasoning mode for your most latency-sensitive use cases. To focus on this newest generation, we are retiring several earlier models effective May 15, 2026 at 12:00 PM PT: grok-4-1-fast-reasoning grok-4-1-fast-non-reasoning grok-4-fast-reasoning grok-4-fast-non-reasoning grok-4-0709 grok-code-fast-1 grok-3 grok-imagine-image-pro Important The deprecated model slugs will continue to work after the retirement date. Requests to these models will automatically redirect as follows: Deprecated reasoning models → grok-4.3 with low reasoning effort Deprecated non-reasoning models → grok-4.3 with none reasoning effort Please note that requests to retired model slugs will be charged at the standard Grok 4.3 pricing of $1.25 / 1m tokens in and $2.50 / 1m tokens out. For additional details, please see our May 15 Model Retirement Guide. Try Grok 4.3 via our Chat Playground. Thank you for building with Grok! So long, and thanks for all the fish, The xAI Team © 2026 X.AI LLC For questions contact support@x.ai