Repository navigation
GPT-6 Astra: repeated "Selected model is at capacity" on Pro 20x — rollout logs show server_overloaded with 0% usage used #43700
Description
What version of the Codex App are you using (From “About Codex” dialog)?
Codex Desktop (bundled codex-cli 0.153.4)
What subscription do you have?
ChatGPT Pro (20x)
What platform is your computer?
Darwin 25.5.0 arm64 (Apple Silicon Mac, macOS)
What issue are you seeing?
Codex Desktop repeatedly terminates turns with:
Selected model is at capacity. Please try a different model.
Model: gpt-6-astra. This has been recurring since Aug 28 and escalated sharply on Sep 7–8.
Local rollout JSONL logs classify all of these terminations as:
"codex_error_info": "server_overloaded"while the rate-limit snapshots in the same logs show the account is nowhere near any limit:
plan_type: pro
used_percent: 0.0
rate_limit_reached_type: null
Measured from my local rollout logs:
| Date (UTC) | Occurrences | Notes |
|---|---|---|
| Aug 28 | 39 banner lines / 30 server_overloaded |
single day |
| Sep 7 | 65 server_overloaded across 5 sessions |
worst session: 27 terminal errors between 07:14–12:27 UTC |
| Sep 8 | 6 server_overloaded between 03:02–04:38 UTC |
today |
Important detail: turns are not rejected upfront — they are admitted, run for a while, then killed mid-flight. In the Sep 7 session the killed turns ran on average 85 s before termination (max 275 s), losing in-progress work on long-running tasks.
Retrying succeeds intermittently, then fails again in bursts. Switching models and starting new sessions does not reliably recover.
What steps can reproduce the bug?
- Sign in with a ChatGPT Pro account (auth via ChatGPT, not API key) in Codex Desktop 0.153.4 on macOS arm64.
- Select
gpt-6-astra. - Start normal coding tasks.
- Turns intermittently terminate with the capacity banner; the rollout log records
codex_error_info: "server_overloaded"whileused_percentstays at 0.0.
Session IDs for server-side correlation:
01a07ab7-4da3-7cf0-b711-638991ba5df0(Sep 7, 27 terminal errors, 07:14–12:27 UTC)01a07ef7-1841-70e0-9344-a8d33ad34714(Sep 8, 6 terminal errors, 03:02–04:38 UTC)
What is the expected behavior?
- The UI message should distinguish server-side capacity / admission-control failures from per-account quota exhaustion — "Selected model is at capacity. Please try a different model." reads like quota guidance, but my quota was at 0% used and switching models does not help.
- Long-running turns should be retried with bounded backoff within the same turn (or resume from a preserved state) instead of being killed mid-flight after minutes of work.
- status.openai.com showed "fully operational" with 100% Codex uptime during these windows; Codex-side regional serving degradation is not reflected there.
Additional information
- The escalation timing coincides with the global paid-usage reset on the evening of Sep 7 (PST); community reports consistently spike right after each reset, which is consistent with a demand surge against fixed serving capacity rather than per-account throttling.
- Community reports suggest Asia/Singapore-serving-region users are hit especially hard, and some users report a VPN region switch mitigates it — which points at regional serving pools.
- This matches the patterns already reported in Repeated "Selected model is at capacity" errors interrupting Codex tasks on ChatGPT Pro #41790, Selected model is at capacity — GPT-5.5, GPT-5.6-Sol, and GPT-6 Astra all failing (only 5.4-mini works), Sep 7 2026 #43398, Repeated "Selected model is at capacity" errors on ChatGPT Pro despite 100% usage remaining; rollout logs show server_overloaded #43663, Multiple GPT-5 / GPT-6 models return "Selected model is at capacity" #43375, gpt-6-astra: websocket closed by server before response.completed on every turn, HTTP fallback 5xx, 'Selected model is at capacity' (0.153.4, ChatGPT Pro) #43517 — filing with my own telemetry/session IDs for correlation; happy for maintainers to merge into an existing tracking issue.
Activity
github-actions commentedon Sep 8, 2026
Potential duplicates detected. Please review them and close your issue if it is a duplicate.
Powered by Codex Action
razdgann commentedon Sep 8, 2026
It would help to expose capacity failures separately from account limits and attach them to the request trace.
Useful fields would include provider error class, retry count, tokens generated before failure, whether the attempt was billable, and the routing decision that followed. Otherwise repeated capacity retries can look like user quota consumption or disappear from cost analysis entirely.
ModelGate is built around preserving those request-level distinctions across providers: https://github.com/razdgann/modelgate-oss
sunjieqing888 commentedon Sep 8, 2026
I would appreciate it if this issue could be addressed promptly. Since a large number of users are affected by the same problem, I kindly request that the official team give it due attention.
FranklinChip commentedon Sep 10, 2026
Update 2026-09-10: Recurring on this account — and now alternating with the separate
invalid_promptfalse-positive reported in #43058.Client unchanged (Codex Desktop, codex-cli 0.153.4, macOS arm64, Pro 20x, ChatGPT sign-in). Model:
gpt-6-astra.Fresh rollout telemetry from today (2 sessions, 6
server_overloadedterminations + 1invalid_prompt):Notable: the two failure modes are now interleaving on the same model/account within minutes of each other. Rate-limit snapshots remain
plan_type: pro,rate_limit_reached_type: nullwith negligible usage consumed. New reports today (#44382) and yesterday (#44113) indicate the capacity wave is recurring after the Sep 7 usage reset.FranklinChip commentedon Sep 10, 2026
Update 2 (2026-09-10, ~03:48 UTC): The impact now extends beyond Codex and beyond a single account.
A separate ChatGPT account (chatgpt.com + ChatGPT desktop app, macOS; account email:
eidelmanjaewoo@mail.com) hit the same error at the same time on GPT-5.6 Sol inside the ChatGPT app itself:Significance for triage:
For context: OpenAI's help center is currently displaying a "service disruption / we are investigating" banner, and the status page shows an active incident (listed under ChatGPT).