GPT-5.5 retires Oct 14 in Codex: what to change for Sume
Codex retires GPT-5.5 on October 14, 2026. A Sume server entry names no model, and Agent Completions only accepts sume-agent, so check your own settings.

Nothing in a Sume setup names a Codex model, so the October 14 retirement does not require a Sume change. GPT-5.5 retires from ChatGPT, ChatGPT Work and Codex on October 14, 2026 (not the OpenAI API). The settings to fix are yours: saved model settings, custom agents, scheduled tasks and scripts that still select gpt-5.5.
Retirement details from the Codex changelog; Sume details from MCP and Agent Completions, read 2026-10-01.
What does the Codex notice say to update?
For Codex with ChatGPT sign-in, choose an available replacement for your plan and client. The notice lists workspace defaults, saved model settings, managed configurations, custom agents, scheduled tasks and scripts that still select gpt-5.5.
Where does a Sume setup mention a model?
| Place | Model field? | Action |
|---|---|---|
Codex MCP entry for https://mcp.sume.com/mcp | None; the entry is a URL plus auth | Nothing to migrate |
Agent Completions model | Only sume-agent; omit it for the same agent | Keep or omit |
| Agent Completions with another id | Rejected with 400 invalid_request | Remove any copied model id |
| Your Codex scripts and schedules | May pin gpt-5.5 | Change them in Codex |
How do I check the connection after switching models?
Start a session on the replacement model and call tools_list, which lists every tool visible to the session, or mcp_health for the endpoint and auth source. If the list is complete, the Sume side is fine. See Codex model picker and the hosted MCP server.
Could a scheduled task still spend money?
A scheduled Codex task that calls a paid Sume tool will keep calling it on the new model, with the same idempotency_key and approval rules. Review those tasks while you are editing the model setting, since a different model may plan different calls.
Sources
Related posts
More in Developers
- gpt-image-2.5-flare-2026-09-08 on Sume: use openai/gpt-image-2.5
OpenAI lists the snapshot gpt-image-2.5-flare-2026-09-08. Sume documents catalog slugs, so send openai/gpt-image-2.5 and check /v1/images/models for ids.
- GPT Image 2.5 Flare: 5 images a minute at Tier 1 vs Sume limits
OpenAI limits Flare to 5 images per minute at Tier 1 and 250 at Tier 5. Sume applies plan concurrency, queue capacity and 429 codes instead. Both tables here.
- GPT Image 2.5 largest square: 2880x2880, not 3000x3000
GPT Image 2.5 caps pixels at 8,294,400. A square tops out at 2880x2880; 3000x3000 (9,000,000) is rejected. OpenAI calls sizes above 2560x1440 experimental.
- GPT Image 2 input_fidelity: omit it; Sume returns 400
OpenAI says to omit input_fidelity for gpt-image-2 because inputs run at high fidelity. Sume lists no such field and rejects unlisted parameters with 400.
Written by Sume