跳转到主要内容

Documentation Index

Fetch the complete documentation index at: https://docs.apiyi.com/llms.txt

Use this file to discover all available pages before exploring further.

2026/5/13 14:32 (UTC+8) · Model Status · OpenAI ⚠️ gpt-image-2 current status: default group saturated, please continue using image2Enterprise — Heads-up: gpt-image-2 on the default group remains saturated (queueing / timeouts on the high side). The enterprise group image2Enterprise is producing images normally. Call logs from this afternoon (5/13 14:29–14:31 UTC+8) show the enterprise group with first-byte latency clustered in the 35–293 second range; non-streaming calls returning normally, prompt / completion token counts healthy. Some default-group requests are being auto-routed to the enterprise group via “Fallback hit: Default → image2Enterprise”.
gpt-image-2 call log 5/13 14:29–14:31: 9 non-streaming calls all routed to image2Enterprise, first-byte 35-293s, 1.2x multiplier, last row shows 'Fallback hit Default → image2Enterprise'
💡 Recommendation: Set your token’s model group directly to image2Enterprise — skip the fallback round-trip (and the default-group queue/failure time). The enterprise group offers high concurrency and stable image generation at a 1.2x multiplier (we don’t profit on it — it’s there to keep capacity flowing). 📖 How to adjust the token group (with screenshots): /en/live/2026-04/image2-enterprise
← Back to Live · 📚 Monthly Archive