Answer: Ultra uses subagents to handle separate parts of a complex task in parallel, and you should choose it when you can divide the work into meaningful parts. Max gives the selected model more time to reason about a single task, for the hardest problems when depth matters more than speed or usage; most tasks do not need Max or Ultra. Status: Confirmed Entry: tool/codex/models/max-vs-ultra-mode Revision: 2 Digest: sha256:2182ea0533eb98a6 (changes whenever the content changes) Timeline: Created: 2026-11-13T03:52:03Z Last changed: 2026-11-18T13:34:17Z (Model support: — → GPT-6 Luna supports reasoning efforts up to **Max**, but not **Ultra**) Last verified: 2026-11-20T13:40:30Z Next check: 2026-11-21T13:40:30Z Check interval: Daily Expected change: No fixed schedule Data: Field | Value | Source | As of When to use Ultra | Choose it when you can divide the work into meaningful parts | [1] | 2026-11-20 Max definition | gives the selected model more time to reason about a single task | [1] | 2026-11-20 When to use Max | for the hardest problems, when depth matters more than speed or usage | [1] | 2026-11-20 General guidance | Most tasks do not need Max or Ultra | [1] | 2026-11-20 Model support | GPT-6 Luna supports reasoning efforts up to **Max**, but not **Ultra** | [1] | 2026-11-20 Sources: [1] OpenAI. "Models". https://learn.chatgpt.com/docs/models.md Role: primary. Accessed: 2026-11-20T13:40:30Z Quote: "Choose it when you can divide the work into meaningful parts." (Supports: When to use Ultra) Quote: "**Max** gives the selected model more time to reason about a single task." (Supports: Max definition) Quote: "Use it for the hardest problems, when depth matters more than speed or usage." (Supports: When to use Max) Quote: "Most tasks do not need Max or Ultra." (Supports: General guidance) Quote: "GPT-6 Luna supports reasoning efforts up to **Max**, but not **Ultra**." (Supports: Model support) Snapshot SHA-256: 2ca411d02af5e803e25d60eedcac752c7a6153c2c6dffafdbd854aff61f4277c (captured 2026-11-18T13:34:16Z) Verification: Method: Automated extraction from [1] with deepseek-flash (JSON output, thinking disabled). Every value was checked by code to appear verbatim in a quoted passage of [1]. On this check the text of [1] was unchanged since 2026-11-18T13:34:17Z, so the values verified then were kept. Cross-check: not-available at 2026-11-20T13:40:30Z. No independent second source is configured for this entry Last successful check: 2026-11-20T13:40:30Z Last attempt: 2026-11-20T13:40:30Z Notes and limitations: Extracted automatically from a single official source (section "Know when to use Max or Ultra"). Related: https://latestn.com/tool/codex/models/recommended-models Which models are recommended for Codex and ChatGPT Work, and what is each model best at? https://latestn.com/tool/codex/models/subagent-model-reasoning-effort How do I choose the model and reasoning effort for Codex subagents? https://latestn.com/tool/codex/permissions-and-security/act-on-security-findings What should I do with the findings from a Codex security change review? https://latestn.com/tool/codex/models/daybreak-model-ci How do I use an approved Daybreak model with Codex Security in CI? https://latestn.com/tool/codex/models/fast-mode-supported-models Which models support Fast mode in Codex? https://latestn.com/tool/codex/models/gateway-recognized-model-names What model names does Codex recognize when routing through a gateway? https://latestn.com/tool/ All entries in AI tools Change log: Time | Event | Change | Source 2026-11-18T13:34:17Z | changed | Model support: — → GPT-6 Luna supports reasoning efforts up to **Max**, but not **Ultra** | [1] 2026-11-18T13:34:17Z | changed | GPT-6 Luna support: supports reasoning efforts up to **Max**, but not **Ultra** → — | [1] 2026-11-18T13:34:17Z | changed | General guidance: Most tasks do not need Max or Ultra. → Most tasks do not need Max or Ultra | [1] 2026-11-18T13:34:17Z | changed | When to use Max: Use it for the hardest problems, when depth matters more than speed or usage. → for the hardest problems, when depth matters more than speed or usage | [1] 2026-11-18T13:34:17Z | changed | When to use Ultra: Choose it when you can divide the work into meaningful parts. → Choose it when you can divide the work into meaningful parts | [1] 2026-11-13T03:52:03Z | added | Entry created | — Changes for this entry as a feed: https://latestn.com/tool/codex/models/max-vs-ultra-mode/latest Challenge this entry: Open challenges: None How: If a value here is wrong or out of date, send what is wrong and a link to evidence. Challenges are queued and checked against the evidence; they do not change the entry by themselves. If the entry changes as a result, the change log says so. Form: https://latestn.com/dispute/tool/codex/models/max-vs-ultra-mode POST: https://latestn.com/dispute/tool/codex/models/max-vs-ultra-mode with claim, evidence_url, and optionally field and contact (form-encoded or JSON) GET: https://latestn.com/dispute/tool/codex/models/max-vs-ultra-mode?claim={what is wrong}&evidence_url={link} How to cite: LatestN. (2026). What is Ultra mode in Codex and when should I use Max or Ultra? Revision 2, last verified 2026-11-20T13:40:30Z. https://latestn.com/tool/codex/models/max-vs-ultra-mode (accessed 2026-11-21T09:59:46Z). @misc{LatestN_tool_codex_models_max_vs_ultra_mode_r2, title = {What is Ultra mode in Codex and when should I use Max or Ultra?}, author = {{LatestN}}, year = {2026}, howpublished = {\url{https://latestn.com/tool/codex/models/max-vs-ultra-mode}}, note = {Revision 2; digest sha256:2182ea0533eb98a6; last verified 2026-11-20T13:40:30Z; accessed 2026-11-21T09:59:46Z} } Reuse: License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/) Attribution: Copying and adapting are allowed, including commercially, if you credit LatestN and link to the entry you used. Each entry ends with a ready-made citation. URL: https://latestn.com/tool/codex/models/max-vs-ultra-mode