GPT-5.2-Codex vs GPT-5.5
GPT-5.2-Codex Verdict
- GPT-5.5 leads on quality: 66.7 vs 60.8.
- GPT-5.2-Codex is stronger in instruction following.
- GPT-5.5 is stronger in agents & tools, coding, composite, knowledge, long context, multimodal, human preference, reasoning.
- GPT-5.2-Codex is 2.3× cheaper ($4.81 vs $11.25 per 1M blended).
- GPT-5.5 streams 3.4× faster (85 vs 25 tokens per second).
| Metric | GPT-5.2-Codex | GPT-5.5 |
|---|---|---|
| BenchLeader Index | 60.8 | 66.7 |
| Agents & tools score | 63.5 | 68.5 |
| Coding score | 55.7 | 57.2 |
| Composite score | 56.8 | 77.8 |
| Instruction following score | 75.1 | 73.6 |
| Knowledge score | 63.1 | 73.9 |
| Long context score | 67.8 | 68.8 |
| Multimodal score | 60.7 | 64.9 |
| Human preference score | – | 67.9 |
| Reasoning score | – | 70.1 |
| Blended price $/M | $4.81 | $11.25 |
| Output speed | 25 tok/s | 85 tok/s |
| Time to first answer | 6.7 s | 58.6 s |
| Context window | 400k | 1.1M |
| Terminal-Bench | 66.5% | 84.7% |
| SimpleBench | – | 69.0% |
| Remote Labor Index | – | 6.3% |
| APEX-Agents | 27.6% | 38.5% |
| FrontierCode | – | 43.0% |
| Epoch Capabilities Index | – | 159.1 |
| LMArena Text | – | 1477 |
| LMArena Hard Prompts | – | 1498 |
| LMArena Coding | – | 1509 |
| LMArena WebDev | 1339 | 1458 |
| LMArena Vision | – | 1296 |
| LMArena Agent | – | 3 |
| LiveBench | 74.0% | – |
| LiveBench Reasoning | 77.7% | – |
| LiveBench Coding | 83.6% | – |
| LiveBench Agentic Coding | 49.4% | – |
| LiveBench Mathematics | 88.8% | – |
| LiveBench Data Analysis | 78.2% | – |
| LiveBench Language | 73.7% | – |
| AA Intelligence Index | 28.5 | 38.6 |
| IFBench | 77.6% | 75.8% |
| AA-LCR | 82.3% | 84.3% |
| MMMU-Pro | 76.3% | 79.9% |
| AA-Omniscience | -2.2 | 20.5 |
| Terminal-Bench Hard | 37.1% | 60.6% |
| GPQA Diamond (AA) | 89.9% | 93.5% |
| Humanity's Last Exam (AA) | 35.7% | 45.8% |
| SciCode (AA) | – | 55.8% |
| τ²-Bench Telecom (AA) | 92.1% | 93.9% |
| LiveCodeBench | 88.0% | – |
| SWE-Bench Pro | 41.0% | – |
| HiL-Bench | – | 39.7% |
| EQ-Bench 4 | – | 1315 |
| Kagi LLM Benchmark | – | 88.8% |
| SWE-bench Verified (bash only) | 72.8% | – |
| SWE-bench Verified (any scaffold) | 72.8% | – |
Data as of 2026-09-10. Best configuration of each model; every score links to its source on the model pages.
GPT-5.2-Codex vs GPT-5.5: questions
- Is GPT-5.2-Codex better than GPT-5.5?
- GPT-5.5 leads on quality: 66.7 vs 60.8. The BenchLeader Index combines every independent quality benchmark; GPT-5.5 is ahead overall as of 2026-09-10, but check the category scores for your use.
- Is GPT-5.2-Codex better than GPT-5.5 for coding?
- GPT-5.5 scores higher in coding (57 vs 56 on the category index, where 50 is average).
- Is GPT-5.2-Codex better than GPT-5.5 for agentic tasks?
- GPT-5.5 scores higher in agentic tasks (69 vs 64 on the category index, where 50 is average).
- Which is cheaper, GPT-5.2-Codex or GPT-5.5?
- GPT-5.2-Codex is cheaper: $4.81 against $11.25 per million tokens, blended at three input tokens per output token.
- Which is faster, GPT-5.2-Codex or GPT-5.5?
- GPT-5.5 streams faster: 85 against 25 output tokens per second.
- Which has the larger context window?
- GPT-5.5 accepts more context: 1.1M against 400k tokens.