OpenAI server_is_overloaded stream errors are not retried

Description

OpenAI-compatible streams can fail with a transient overload event instead of normal text:

{"type":"error","error":{"type":"service_unavailable_error","code":"server_is_overloaded","message":"Our servers are currently overloaded. Please try again later."}}

When that shape is surfaced as a stream error, it should enter the existing session retry flow instead of stopping the turn as a final error.

Plugins

None required to reproduce the underlying provider error shape.

OpenCode version

1.14.33

Steps to reproduce

Use an OpenAI-compatible model during a provider overload window.
The provider returns the stream error above.
The session stops instead of showing the existing retry countdown/status.

Screenshot and/or share link

N/A

Operating System

macOS

Terminal

Any

This issue doesn't fully meet our contributing guidelines.

What needs to be fixed:

This issue appears to have been generated by an AI tool (aidevops.sh, as noted in the footer). AI-generated walls of text are not acceptable — please rewrite this in your own words, describing the actual problem you experienced and why you need this feature.
Feature requests must start with [FEATURE]: in the title.

Please edit this issue to address the above within 2 hours, or it will be automatically closed.

This issue might be a duplicate of existing issues. Please check:

Codex server_is_overloaded stream errors are not retried #25730: Codex server_is_overloaded stream errors are not retried
502 provider_unavailable errors from OpenRouter are not retried, causing subagent/session aborts #22448: 502 provider_unavailable errors from OpenRouter are not retried, causing subagent/session aborts
Anthropic overloaded_error is not retried, halts session #20384: Anthropic overloaded_error is not retried, halts session
Session retry loops forever — no max attempt cap or configurable limit #25733: Session retry loops forever — no max attempt cap or configurable limit

If you believe this was flagged incorrectly, please let a maintainer know.

Thanks for updating the issue! It now meets the contributing guidelines and includes all the necessary information. We appreciate you taking the time to provide a clear description, reproduction steps, and system details.

test

Thanks for the automated guidance. I updated the issue and PR to match the repo templates and removed the generated footer/long-form wording.

I also checked the related work and saw #25728 covers the same server_is_overloaded stream shape from the Codex side. I pushed one small adjustment here so this branch also routes the stream error through parseStreamError, not just the later retry classifier.

If maintainers prefer #25728, I can close this PR as a duplicate. Otherwise this one is now the same narrow bug fix from the OpenAI-compatible stream angle.

Please prioritise.

I am suprised. This almost makes OpenAI service unusable at all with OpenCode

Yes please fix this

service_unavailable_error: Our servers are currently overloaded. Please try again later.

This needs to be fixed. Also I am shocked we all have this issue yet codex status shows that everything is absolutely fine. What am I even paying my money for at this point...

Yup this started happening with GPT5.5 on pro subs from yday. Probably needs to be retried.

Fixed, but on dev. I built from dev and it works. Feel free to release.

Related fix in #30323 — covers server_is_overloaded and other OpenAI/Codex stream envelopes in parseStreamError + retryable().

The issue is still valid. I can see related PR with fix is closed now...

This issue has been relentless today, had to switch to pi to get work done. Happening on v1 latest and v2 latest. I suspect it's only reproducible under certain conditions at OpenAI.

Checked the reported type: "error" stream frame with nested error.code: "server_is_overloaded" against the current V1 (dev) and V2 (v2) code:

V1 routes that frame through parseStreamError as retryable; its session retry path also recognizes overload messages. The broader retry handling landed in refactor(opencode): simplify retry error matching #40694 and fix(opencode): expand retryable error patterns #40707, so fix: retry OpenAI overload errors #25886 is not needed for this specific frame.
V2's Responses stream handler classifies the exact frame as ProviderInternal, which session retry treats as retryable. I checked that exact payload against the V2 classifier.

Closing for the specific payload in this issue. The August report of failures on both versions may involve another error shape or retry condition; if it still occurs, please provide a fresh version and the actual stream/error payload so that path can be investigated separately.

Navigation Menu

Repository navigation

OpenAI server_is_overloaded stream errors are not retried #25884

Description

Plugins

OpenCode version

Steps to reproduce

Screenshot and/or share link

Operating System

Terminal

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

OpenAI server_is_overloaded stream errors are not retried #25884

Description

Description

Plugins

OpenCode version

Steps to reproduce

Screenshot and/or share link

Operating System

Terminal

Activity

github-actions commented on May 5, 2026

github-actions commented on May 5, 2026

github-actions commented on May 5, 2026

johnwaldo commented on May 5, 2026

samiralibabic commented on May 5, 2026

EnderHane commented on May 6, 2026

MichaelFisher1997 commented on May 6, 2026

doodlezz commented on May 6, 2026

AleksQuantum commented on May 6, 2026

Techie5879 commented on May 6, 2026

samiralibabic commented on May 6, 2026

spark4862 commented on Jun 2, 2026

dawidgora commented on Jul 17, 2026

fareesh commented on Aug 11, 2026

opencode-agent commented on Oct 1, 2026

Metadata

Metadata

Assignees

Labels

Type

Projects

Milestone

Relationships

Development

Issue actions