GPT-5-Codex

Tool callingStructured outputReasoningImage input

Launch post Website Docs

Coding-optimized GPT model for repository edits, reviews, and agentic software work

Context: 400K
Released: 1y ago

Sources:models.dev

Benchmarks

DeepSWE is the primary coding benchmark where an exact model evaluation exists. Other sourced evaluations appear below; missing scores are not zero.

Not evaluated in DeepSWE 1.1 as of 2026-09-22. Results are only shown for exact model versions in the official dataset. Official leaderboard.

Other published benchmarks

Evaluation	Reported score and source
Terminal Bench	44.3% Terminal-Bench v2 Leaderboard · gpt-5-codex Agent: Codex CLI Terminal-Bench v2 Leaderboard Original-source terms apply
Terminal Bench	41.3% Terminal-Bench v2 Leaderboard · gpt-5-codex Agent: Mini-SWE-Agent Terminal-Bench v2 Leaderboard Original-source terms apply
Terminal Bench	43.4% Terminal-Bench v2 Leaderboard · gpt-5-codex Agent: Terminus 2 Terminal-Bench v2 Leaderboard Original-source terms apply
Terminal Bench	44.3% https://www.tbench.ai/leaderboard/terminal-bench/2.0 · gpt-5-codex Agent: Codex CLI https://www.tbench.ai/leaderboard/terminal-bench/2.0 Original-source terms apply
Terminal Bench	41.3% https://www.tbench.ai/leaderboard/terminal-bench/2.0 · gpt-5-codex Agent: Mini-SWE-Agent https://www.tbench.ai/leaderboard/terminal-bench/2.0 Original-source terms apply
Terminal Bench	43.4% https://www.tbench.ai/leaderboard/terminal-bench/2.0 · gpt-5-codex Agent: Terminus 2 https://www.tbench.ai/leaderboard/terminal-bench/2.0 Original-source terms apply

Benchmark compilation: Epoch AI, Capabilities & benchmarking (CC BY attribution). Scores are converted to percentages and model names matched to the Vertical Tabs catalog. Individual evaluations are credited to their original sources. External results may carry separate licences and evaluation configurations; do not interpret scores from different harnesses as directly comparable.

Specifications

Developer: OpenAI
Model family: GPT
Type: Language model
Predecessor: —
Successor: —
Released: September 15, 2025
Model ID: openai/gpt-5-codex
Openness: Closed
License: —
Input modalities: text, image
Output modalities: text
Modality: text+image->text
Max output tokens: 128K
Knowledge cutoff: 2024-09-30
Hugging Face: —
Source freshness: Metadata synced Oct 8, 2026, 10:17 PM UTC

Additional source evidence

Source model: openai/gpt-5-codex
Source family: gpt-codex
Last updated: 2025-09-15
Knowledge cutoff: 2024-09-30
Model context: 400K
Maximum input: 272K
Model output limit: 128K
Open weights: No
File attachments: No
Reasoning: Yes
Tool calling: Yes
Structured output: Yes
Temperature control: No

Published benchmarks

Benchmark	Score	Evaluation details	Evidence
Artificial Analysis Coding Index	38.9 · index	2026-06-01	Source
SciCode	40.9 · percent correct	2026-06-01	Source
Terminal-Bench Hard	37.9 · success rate	2026-06-01	Source

Complete source record (JSON)

{
  "id": "openai/gpt-5-codex",
  "name": "GPT-5-Codex",
  "description": "Coding-optimized GPT model for repository edits, reviews, and agentic software work",
  "family": "gpt-codex",
  "last_updated": "2025-09-15",
  "knowledge": "2024-09-30",
  "attachment": false,
  "reasoning": true,
  "tool_call": true,
  "structured_output": true,
  "temperature": false,
  "modalities": {
    "input": [
      "text",
      "image"
    ],
    "output": [
      "text"
    ]
  },
  "limit": {
    "context": 400000,
    "input": 272000,
    "output": 128000
  },
  "benchmarks": [
    {
      "name": "Artificial Analysis Coding Index",
      "score": 38.9,
      "source": "https://openrouter.ai/openai/gpt-5-codex/benchmarks",
      "metric": "index",
      "date": "2026-06-01"
    },
    {
      "name": "SciCode",
      "score": 40.9,
      "source": "https://openrouter.ai/openai/gpt-5-codex/benchmarks",
      "metric": "percent correct",
      "date": "2026-06-01"
    },
    {
      "name": "Terminal-Bench Hard",
      "score": 37.9,
      "source": "https://openrouter.ai/openai/gpt-5-codex/benchmarks",
      "metric": "success rate",
      "date": "2026-06-01"
    }
  ],
  "release_date": "2025-09-15",
  "open_weights": false
}

Infrastructure

Provider availability

All providers

Provider endpoint data will appear after this model's first OpenRouter endpoint sync.

Text provenance

EU · product-scoped

OpenAI textGrain watermarking

In the EU, ChatGPT-generated text uses textGrain. OpenAI says API customers can opt in globally for select models, so watermarking depends on the product surface and configuration.

Official explanation