Introducing upgrades to Codex (OpenAI) Sep 15, 2025
GPT-5-Codex
GPT-5-Codex places in the top three on Terminal-Bench.
14
Benchmarks with a score
0
Ranked first, out of five or more models
1
In the top three
OpenAI Sep 2025
GPT-5-Codex
Closed weights
No Epoch Capabilities Index score yet
- Terminal-Bench 2.0 44.3% #21 /43
- Terminal-Bench 44.3% #3 /18
- BullshitBench V2 39% #64 /129
Where it ranks
Its best score on each benchmark, and where that ranks among every model with a score there. We rank each model once, at its best setting.
Knowledge and language
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| BullshitBench | 30.9% | #47 of 107 | Official leaderboard | |
| BullshitBench V2 | 39% | #64 of 129 | Official leaderboard |
Coding
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| SWE-bench Verified | 74.5% | #49 of 155 | Lab model card | |
| WeirdML | 54.5% | #40 of 112 | Official leaderboard | |
| MLE-bench | 48.4% | #5 of 12 | Official leaderboard | |
| IOI (Vals) | 9.75% | #33 of 55 | Official leaderboard | |
| DesignArena Agentic Overall | 1,043 | #62 of 74 | Official leaderboard | |
| Code refactoring tasks (OpenAI) | 51.3% | #1 of 1 | Lab model card |
Agents and tools
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| SnakeBench | 33 | #15 of 188 | Official leaderboard | |
| Terminal-Bench | 44.3% | #3 of 18 | tbench.ai, via Epoch AI | |
| CompileBench | 93.3% | #4 of 20 | Official leaderboard | |
| Terminal-Bench 2.0 | 44.3% | #21 of 43 | Official leaderboard |
Safety and security
| Benchmark | Score | Rank | Source | |
|---|---|---|---|---|
| Malware Refusals (Golden Set) | 100% | #1 of 3 | Lab model card | |
| Coding prompt injections | 98% | #3 of 4 | Lab model card |
Its launch post
The benchmarks the lab chose to headline, with the scores it gave.
Its system card
Every result in the longer documents the lab published with it. 4 of them aren't in the launch post.
Addendum to GPT-5 system card: GPT-5-Codex Sep 15, 2025