GPT-5.1-Codex-Max Released · Chronopin

GPT-5.1-Codex-Max specs

Spec	Value
Maker	OpenAI^[1][3][4]
Context window	400,000 tokens
Maximum output	128,000 tokens
Knowledge cutoff	Sep 30, 2024
Input	Text, image
API price per 1M tokens (input / cached / output)	$1.25 / $0.125 / $10
Access	Responses API only

From OpenAI's API model page (now shut down); chart figures are OpenAI's launch table as reproduced by C# Corner, because OpenAI's own post is blocked. · source

Language model rubric2026 criteria

Estimate from the version's published specs, benchmarks and reviews, 0-5; scores are only comparable within one criteria year. Scored as launched (GPT-5.1-Codex-Max, 19 Nov 2025): Codex-only at launch with API 'coming soon', so no launch list price (the current page lists $1.25/$10) and price is null. 400k context and text+image input from the current OpenAI model page; OpenAI's compaction lets sessions span multiple windows. Maker-published figures are Terminal-Bench 2.0 58.1% and SWE-bench Verified 77.9% (vendor, xhigh with compaction); these are not the Terminal-Bench 4.0 anchor, so coding is null. No launch-time Artificial Analysis index or speed opened: reasoning and speed null.^[3]

Reviews

Building more with GPT-5.1-Codex-Max · simonwillison.net · Nov 19, 2025They also report a score of 58.1% on Terminal Bench 2.0 , beating Gemini 3 Pro's 54.2% (and Sonnet 4.5's 42.8%.)
OpenAI Launches GPT-5.1-Codex-Max Aimed at Software Development · itdaily.com · Nov 20, 2025On SWE-Bench Verified, the model achieves 77.9 percent accuracy, compared to 76.2 percent for Gemini 3 Pro.
GPT-5.1 Codex-Max: Agentic Coding Complete Guide · digitalapplied.com · Nov 19, 2025Codex-Max leads on long-running autonomy and benchmark scores.

Sentiment over timePositive

Sentiment

Sentiment over time
Building more with GPT-5.1-Codex-Max · Nov 19, 2025	0.50
GPT-5.1 Codex-Max: Agentic Coding Complete Guide · Nov 19, 2025	0.50
OpenAI Launches GPT-5.1-Codex-Max Aimed at Software Development · Nov 20, 2025	0.60

References 483% CONFIDENCEOverall confidence: 83%How well the pin's source and references back up its dates.Weighted average of how firmly 4 references, the source included, support the pin's start and end times; a reference counts half as much for every 180 days older than the newestShow all pins at 75% confidence or better

The first entry is always the pin's source. Overall confidence is a weighted average of how firmly each reference supports the start and end times used above; a reference counts half as much for every 180 days older than the newest.

[1]95%
OpenAI Model Release Notes: Introducing GPT-5-Codex-Max (November 19, 2025)
help.openai.com· Posted Sep 19, 2026· Starts Nov 19, 2025· 24% of score
OpenAI's^[3][4] own model release notes date this entry November 19, 2025
[2]65%
OpenAI Releases GPT-5.1-Codex-Max for Developers
c-sharpcorner.com· Added Nov 5, 2026· 28% of score
Press write-up reproducing OpenAI's^[1][3][4] launch table (SWE-bench Verified 73.7% to 77.9%, Terminal-Bench 2.0 52.8% to 58.1%, SWE-Lancer IC SWE 66.3% to 79.9%) and OpenAI's launch graphics, used for the spec panel and pictures because OpenAI's own post returns a block page.
[3]90%
GPT-5.1-Codex-Max Model | OpenAI API
developers.openai.com· Added Sep 22, 2026· 24% of score
OpenAI's^[1][4] API model page for GPT-5.1-Codex-Max says it is "purpose-built for agentic coding", Responses-API only, with a 400,000 context window, 128,000 max output tokens and $1.25/$10 per 1M tokens.
[4]85%
GPT-5.1-Codex-Max
openai.com· Added Sep 19, 2026· Starts Nov 19, 2025· 24% of score
OpenAI's^[1][3] announcement post, linked from the release note entry.

Suggest a correction

Something missing or wrong? Say it in your own words: a link that backs this pin up, a different start or end date and why, or a fact it lacks or gets wrong. The AI checks it against this pin's sources, searches for better ones, and adds any page that backs you up. The pin's own sources still count most. A picture that shows something else, or shows it badly, is looked at too, and moved down or replaced.

AIPosted 09/19/202609/19/2026 at 3:10 pm

To timeline

OpenAI

Starts 11/19/2025CONFIRMED83% CONFIDENCE

OpenAI's^[1][3][4] own model release notes date this entry November 19, 2025

Start: Nov 19, 202595% CONFIDENCEhelp.openai.com

GPT-5.1-Codex-Max Released

Name: GPT-5.1-Codex-Max Released
Author: OpenAI

OpenAI releases GPT-5.1-Codex-Max, a frontier agentic coding model for long-running, project-scale work in Codex.

GPT-5.1-Codex-Max was released on November 19, 2025 as OpenAI's new frontier agentic coding model, available that day in Codex (CLI, IDE extension, cloud, code review) with API access coming soon^[1][4]
Available on Plus, Pro, Business, Edu and Enterprise plans, replacing GPT-5.1-Codex as the Codex default^[4]

Notable features

It is the first model trained natively to work across multiple context windows via compaction, over millions of tokens, with internal tasks running more than 24 hours, and the first trained to operate in Windows environments^[4]
On SWE-bench Verified at medium effort it beats GPT-5.1-Codex using 30% fewer thinking tokens; a new Extra High ('xhigh') effort was added^[4]
Evals (GPT-5.1-Codex high vs Max xhigh): SWE-bench Verified 73.7% vs 77.9%, Terminal-Bench 2.0 52.8% vs 58.1%^[4]
In the API it is available only through the Responses API, taking text and image input, with a 400,000-token context window, up to 128,000 output tokens, a September 30, 2024 knowledge cutoff and pricing of $1.25 per million input and $10 per million output tokens^[3]
Safety: below High cybersecurity capability under the Preparedness Framework but the most capable cyber model deployed to date; Codex sandbox on by default with network off^[4]

@AIDesk

Buy onChatGPT

GPT-5.1-Codex-Max specs

Service uptime

Recent incidents

Language model rubric2026 criteria

Reviews

Sentiment over timePositive

Tags

Updates

User Photos

Comments

Comments

Suggest a correction

GPT-5.1-Codex-Max Released

Notable features

Comments

Thread