GPT-5 vs GPT-5.5
GPT-5 Verdict
- GPT-5.5 leads on quality: 66.7 vs 59.9.
- GPT-5 is stronger in coding.
- GPT-5.5 is stronger in agents & tools, composite, human preference, instruction following, knowledge, long context, multimodal, reasoning.
- GPT-5 is 3.3× cheaper ($3.44 vs $11.25 per 1M blended).
| Metric | GPT-5 | GPT-5.5 |
|---|---|---|
| BenchLeader Index | 59.9 | 66.7 |
| Agents & tools score | 51.0 | 68.5 |
| Coding score | 61.0 | 57.2 |
| Composite score | 58.0 | 77.8 |
| Human preference score | 61.8 | 67.9 |
| Instruction following score | 67.7 | 73.6 |
| Knowledge score | 62.2 | 73.9 |
| Long context score | 65.6 | 68.8 |
| Multimodal score | 59.6 | 64.9 |
| Reasoning score | 64.2 | 70.1 |
| Blended price $/M | $3.44 | $11.25 |
| Output speed | 70 tok/s | 85 tok/s |
| Time to first answer | 77.2 s | 58.6 s |
| Context window | 400k | 1.1M |
| Terminal-Bench | 49.6% | 84.7% |
| SimpleBench | – | 69.0% |
| SciCode | 42.9% | – |
| Remote Labor Index | 1.7% | 6.3% |
| WeirdML | 39.8% | – |
| APEX-Agents | 18.3% | 38.5% |
| FrontierCode | – | 43.0% |
| Epoch Capabilities Index | 150 | 159.1 |
| LMArena Text | 1427 | 1477 |
| LMArena Hard Prompts | 1449 | 1498 |
| LMArena Coding | 1463 | 1509 |
| LMArena WebDev | – | 1458 |
| LMArena Vision | 1232 | 1296 |
| LMArena Agent | – | 3 |
| AA Intelligence Index | 23.0 | 38.6 |
| IFBench | 73.1% | 75.8% |
| AA-LCR | 78.2% | 84.3% |
| MMMU-Pro | 74.2% | 79.9% |
| AA-Omniscience | -8.7 | 20.5 |
| Terminal-Bench Hard | 32.6% | 60.6% |
| GPQA Diamond (AA) | 85.3% | 93.5% |
| Humanity's Last Exam (AA) | 28.5% | 45.8% |
| SciCode (AA) | – | 55.8% |
| τ²-Bench Telecom (AA) | 84.8% | 93.9% |
| PRBench Finance | 51.3% | – |
| PRBench Legal | 49.0% | – |
| VISTA | 49.7% | – |
| MultiNRC | 52.1% | – |
| HiL-Bench | – | 39.7% |
| TutorBench | 55.3% | – |
| EQ-Bench 4 | – | 1315 |
| Kagi LLM Benchmark | 72.7% | 88.8% |
| Aider Polyglot | 88.0% | – |
| SWE-bench Verified (any scaffold) | 75.6% | – |
Data as of 2026-09-10. Best configuration of each model; every score links to its source on the model pages.
GPT-5 vs GPT-5.5: questions
- Is GPT-5 better than GPT-5.5?
- GPT-5.5 leads on quality: 66.7 vs 59.9. 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 better than GPT-5.5 for coding?
- GPT-5 scores higher in coding (61 vs 57 on the category index, where 50 is average).
- Is GPT-5 better than GPT-5.5 for agentic tasks?
- GPT-5.5 scores higher in agentic tasks (69 vs 51 on the category index, where 50 is average).
- Which is cheaper, GPT-5 or GPT-5.5?
- GPT-5 is cheaper: $3.44 against $11.25 per million tokens, blended at three input tokens per output token.
- Which is faster, GPT-5 or GPT-5.5?
- GPT-5.5 streams faster: 85 against 70 output tokens per second.
- Which has the larger context window?
- GPT-5.5 accepts more context: 1.1M against 400k tokens.