Codex long-running tasks repeatedly hit server_overloaded despite 70% usage left
Lambo-dash · ghdev · 2026-10-09
A Windows user reports that OpenAI Codex Desktop repeatedly throws "Selected model is at capacity" errors during long-running multi-step tasks, even though the usage dashboard shows about 70% remaining.
- Affected models include GPT-6.1 Sol and GPT-6 Astra (GPT-6 Sol also seen while switching)
- Switching models lets the task resume briefly, but the error recurs quickly, with reconnect attempts hitting 5/5 in the UI
- The issue worsened noticeably on 2026-10-09 (UTC+08:00); the previous day was relatively normal
- A read-only PowerShell search of session .l logs found 41 matches for serveroverloaded and 41 for "at capacity" (the author notes these may reflect the same events)
- Local pairing works fine; the problem is specific to remote long-task sessions
More from coding & agent
- Decisions API hits GA as agents automate cloud video editing via feedback loops — dkundel · 2026-10-23
- Swarms API changelog: 8 months of streaming, multi-agent orchestration, and dynamic model pricing — KyeGomezB · 2026-10-23
- A curated library of 50+ free AI agent resources: videos, repos, one Google Doc — ifioknkem · 2026-10-23
- Vercel adds no-account deploys: drag and drop a file to get a live URL — steren · 2026-10-23
- Dev runs 5 Claude Code sessions that coordinate via MCP Agent Mail to root his network cameras — doodlestein · 2026-10-23
- Vercel ships 'vercel deploy --temporary' for ephemeral agent-built deployments — cramforce · 2026-10-23