GPT-5.1 Codex
GPT-5.1 Codex is OpenAI's agentic-coding variant of GPT-5.1, built for Codex harnesses, available only via the Responses API with apply_patch and shell tooling.
Model Overview
Capabilities, design details, and architectural traits
GPT-5.1 Codex - GPT-5.1 Tuned for Agentic Coding in Codex
GPT-5.1 Codex is OpenAI's coding-focused variant of GPT-5.1, built to run inside Codex or similar agent harnesses rather than as a general chat model. It is offered only through the Responses API, where the underlying snapshot is updated over time instead of staying fixed at one release.
| Trait | Detail |
|---|---|
| Defining purpose | A version of GPT-5.1 optimized for agentic coding tasks inside Codex or Codex-like harnesses, not general-purpose use. |
| Deployment model | Available only through the Responses API, with the model snapshot regularly updated rather than locked at launch. |
| Companion tools | Built to work with apply_patch, a freeform structured-diff tool for file edits without JSON escaping, and a shell tool for running commands directly. |
Released Alongside a Smaller Sibling
OpenAI released GPT-5.1 Codex Mini on the same day, positioned as a more cost-efficient option for the same coding and agentic workloads. The pairing reflects a two-tier approach: one Codex model tuned for capability, one tuned for cost, both built on the same GPT-5.1 base.
Benchmark Performance
Independent evaluations · Artificial Analysis
Accuracy & Capability Details
Compare Models Side-by-Side
Evaluate specifications, pricing, and independent benchmark indices
| Model Details | |||
|---|---|---|---|
| General Info | |||
| Provider | OpenAI | Anthropic | Anthropic |
| Release Date | November 13, 2025 | September 22, 2026 | September 28, 2026 |
| Knowledge Cutoff | - | - | Jun 2026 |
| Context & Limits | |||
| Context Window | 400K | 1M Best Context Window | 1M Best Context Window |
| Pricing (per 1M tokens) | |||
| Input Pricing | $1.25 Best Input Pricing | $4 | $2 |
| Output Pricing | $10 Best Output Pricing | $20 | $10 Best Output Pricing |
| Modalities | |||
| Inputs | textimage | textimagefile | textimagefile |
| Outputs | text | text | text |
| Benchmarks (0-100) | |||
| Intelligence Index | 23.7 | 57.6 Best Intelligence Index | 56.0 |
| Coding Index | - | - | - |
| Agentic Index | - | - | - |
Humanity's Last Exam
Extremely difficult logical reasoning and knowledge.
Long Context Reasoning
Logical reasoning over long context windows.
SciCode Benchmark
Scientific coding and mathematical modeling.
Independent evaluation data provided by Artificial Analysis. To view the latest benchmarks and full details, visit their official site.
Explore more from OpenAI
Other models by OpenAI
Top AI Models
Leading alternatives by intelligence score
