xAI now lets requests set `service_tier: "priority"` on text, image, and video inference endpoints.

Priority is now request-level, with billing applied only when the priority tier is actually used.

The response reports the applied tier, so builders can verify scheduling priority per call across supported inference endpoints.

• Jun 16, 2026OpenAI modelcard removed API docs navigation text from the captured page.
• Jun 16, 2026Anthropic lists Claude Fable 5 and Claude Mythos 5 on its model overview.
• Jun 16, 2026Google changed the Gemini Developer API pricing page.

This is the world we build in. It moves this fast. We move with it.