Model model_at_capacity api proxy

Selected model is at capacity

This message usually means the selected Codex model or model route has no available capacity right now. It is not primarily an API-key error or local configuration failure; switch models, wait briefly, or ask your proxy provider to check model-route capacity.

Snapshot

Error: Selected model is at capacity. Please try a different model. Treat it as model-level capacity or routing congestion first: switch models, wait briefly, reduce context, and ask your proxy provider to check account-pool and model-group availability.

#selected model is at capacity #codex selected model is at capacity #codex model at capacity #codex please try a different model #openai codex model capacity

What This Error Means

Selected model is at capacity. Please try a different model. is mainly about the selected model. The request is not primarily failing at authentication; the selected model, model group, or upstream route cannot currently allocate capacity.

If you are using the official Codex login flow, this may be temporary OpenAI-side model demand. If you are using a proxy, Sub2API, CPA, NewAPI, or a team gateway, it may also mean the proxy’s account pool, model mapping, or route has no healthy upstream available.

Do not reduce this to “Codex is broken.” The sharper diagnosis is: this model route cannot schedule resources right now.

The OpenAI Developer Community had a matching report on April 23, 2026: a user said gpt-5.4 consistently showed this message, and later replies reported long-running workflows being interrupted in Codex Desktop on macOS and Windows. That pattern fits Codex client behavior, model capacity, and streaming stability more than a standard HTTP 429.

In some Sub2API setups, users observed that the upstream failure is closer to Our servers are currently overloaded, and the proxy layer then surfaces it as the Codex at-capacity message. The operational problem is not only the failure itself; Codex may not automatically continue after this message, so long coding tasks get interrupted.

Common Causes

  1. The selected Codex model is under high demand, and the service is asking you to choose another model.
  2. Your proxy maps the selected model to an upstream account pool that is full, rate limited, or cooling down.
  3. Your API key group has no available supplier for that Codex model, or the model is only enabled for some plans.
  4. The current task has too much context or too many tool calls, making a high-cost model more likely to hit capacity controls.
  5. Multiple Codex sessions are sharing the same account, key, or provider route at the same time.

How To Fix It Quickly

  1. Switch to another available model first. Do not keep hammering the same peak-demand Codex, pro, or high-reasoning model.
  2. Wait 1 to 5 minutes before retrying. Capacity errors are often temporary, and immediate repeated retries rarely help.
  3. Reduce context pressure: start a fresh session, attach fewer files, narrow the requested change, or split the work into smaller steps.
  4. If you use an API proxy, switch model groups, backup routes, or providers, and ask support whether this model has schedulable upstream accounts.
  5. If you use the official Codex App or VS Code extension, keep the screenshot, time, model name, and request ID, then report it through Codex feedback or GitHub issues.
  6. If every model fails, check the OpenAI status page, your proxy status page, balance, plan permissions, and Base URL configuration.

Sub2API Mitigation

If you run or administer Sub2API, you can try an error passthrough rule under account management. Match upstream messages containing Our servers are currently overloaded or at capacity, then rewrite them to an error text that your client treats as retryable, such as upstream failed.

This reduces long-task interruptions; it does not create more model capacity. Keep retries bounded, and watch usage, failure rate, and loops. Hiding every capacity error can blind you to a real capacity problem and burn balance on useless retries.

How It Differs From 429 And 503

429 is more about request rate, quota, or throttling. 503 is more about temporary service unavailability or no available upstream account.

Selected model is at capacity is not 429. It more specifically points to the selected model or upstream model route having no available capacity, sometimes underneath an upstream overloaded response or streaming interruption. Your first move should be switching model, switching model route, or using bounded retry, not regenerating keys, reinstalling Codex, or migrating the entire provider.

One hard truth: if you use a proxy, this message does not automatically prove official Codex is unstable. It may simply mean the proxy oversold a popular model route. Run a minimal request and switch-model test before blaming the whole stack.

Sources

Related errors

Claude Code / Codex 503 No available accounts Do not rotate keys first. This error usually points to an unavailable account pool, provider, group, or model route. Test a minimal request, confirm the model and group, then decide whether to wait, reduce context, change model, or switch provider. Error 429 Too Many Requests for AI API proxies Error: exceeded retry limit, last status: 429 Too Many Requests, request id: <request-id>. With an AI API proxy, repeated 429s usually mean the upstream account pool is rate limited, cooling down, or out of usable quota. Ask the provider to switch account, route, or model first. Error 401 Unauthorized for AI API proxies A 401 is an authentication failure. Check which key is loaded, which Base URL receives the request, whether the header format is correct, and whether the key is still active for that provider.

Related guides

Quickstart: connect Claude Code, Codex, or Cursor to an AI API proxy Before connecting Claude Code, Codex, or Cursor to an AI API proxy, confirm the tool, protocol compatibility, Base URL, API key, and model name. Codex OpenAI-compatible proxy Base URL setup guide Install the OpenAI Codex CLI and configure API keys, Base URL, config.toml, and provider switching for an OpenAI-compatible proxy. Use CC Switch to manage Claude Code and Codex proxy configs Use CC Switch to manage API proxy providers, Base URLs, API keys, config switching, and sync for Claude Code, Codex, and other AI coding CLIs.

Related topics

Codex Proxy Service Comparison Compare Codex proxy services and AI API providers by Codex support, Base URL setup, Responses API notes, pricing, payment methods, and verification signals. OpenAI API Proxy Providers Compare OpenAI API proxy and OpenAI-compatible providers by Base URL support, models, pricing, payment methods, and public verification signals. AI API Directory Browse an AI API directory of OpenAI-compatible, Claude-compatible, and multi-model proxy providers with pricing, payment, model, and verification fields.

FAQ

Common questions

Is this an API key error?

Usually no. API-key problems more often return 401 or 403. Selected model is at capacity points more directly to model capacity, routing, or an upstream account pool that cannot serve the chosen model.

Should I keep retrying the same model?

No. Repeated retries can amplify a capacity problem. Wait briefly or switch to another available model. If you use an API proxy, ask the provider whether that model route is currently full.

Why does normal chat work while Codex says the model is full?

Codex tasks often carry longer context and tool calls, and they may use a dedicated Codex model or model group. A normal chat model being available does not prove the selected Codex route has capacity.