Codex hit by cross-model "at capacity" errors, routing failure suspected
LiangDazhu · ghdev · 2026-09-07
A developer filed an issue on openai/codex reporting repeated "Selected model is at capacity" errors across multiple models, including GPT-6 Astra and GPT-5.6 variants.
Key points:
- The error affects several models simultaneously; switching models often triggers the same error and retries don't reliably recover;
- The author suspects a Codex backend capacity or request-routing problem rather than any single model being overloaded;
- Suggested fixes: auto-routing to available backends, distinguishing model capacity vs. account throttling vs. regional capacity vs. serving failures, and surfacing request IDs for diagnostics.
More from Models
- Sol61 flunks instruction-following: skips mandated research even when explicitly told — lucasmeijer · 2026-10-26
- Kimi K3.1 Pro tops MathArena open models, closing frontier gap to about 6 months — teortaxesTex · 2026-10-26
- Ruff creator Charlie Marsh: frontier models find infinite bugs in any software — charliermarsh · 2026-10-26
- Dev testing one skill across Opus 5.5 and Sol 6.1 finds instruction following 'all over the map' — lucasmeijer · 2026-10-26
- supermemory unveils learner-1, claiming continual learning for any model — julianweisser · 2026-10-26
- Anthropic says Opus 5.5 cuts cost per task by ~40% vs Opus 5 — taufiqintech · 2026-10-26