Model comparison
GPT-4 vs GPT-5.1
GPT-5.1 is the stronger model overall, scoring 49.0 to 29.1 on the Noometry Index.
Last verified . 26 shared benchmarks.
Summary
- They share 26 benchmarks with published results for both. GPT-4 scores higher in 0 categories and GPT-5.1 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.1 leads 52.2 to 10.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.1% for GPT-4 and 88.6% for GPT-5.1.
- GPT-5.1 is cheaper at $1.25 / $10 per million input/output tokens, against $30 / $60 for GPT-4.
- GPT-5.1 accepts more context: 400K tokens versus 8K.
Side by side
| GPT-4 | GPT-5.1 | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 29.1 | 49.0 |
| Released | 2023-03-14 | 2025-11-13 |
| Weights | Proprietary | Proprietary |
| Context window | 8K | 400K |
| Max output | 8K | 128K |
| Input $ / M tokens | $30 | $1.25 |
| Output $ / M tokens | $60 | $10 |
| Results tracked | 38 | 63 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.1 leads
GPT-4: 31.6 (#283), GPT-5.1: 46.4 (#66)
| Benchmark | GPT-4 | GPT-5.1 |
|---|---|---|
| WeirdML | 12.4% | 60.8% |
| LMArena Coding | 1254 | 1454 |
| SWE-bench Verified | — | 68% |
| SWE-bench Verified (bash only) | — | 66% |
| LMArena WebDev | — | 1395 |
| SciCode | — | 43.3% |
| GSO | — | 13.7% |
| BigCodeBench Instruct | 46% | — |
| LiveBench Coding | — | 72.5% |
| BigCodeBench Complete | 57.2% | — |
| ALE-Bench | — | 1,192 |
| HumanEval+ | 79.3% | — |
Agentic & Tool Use Not comparable
GPT-4: —, GPT-5.1: 32.7 (#60)
| Benchmark | GPT-4 | GPT-5.1 |
|---|---|---|
| Terminal-Bench | — | 47.6% |
| DeepResearch Bench | — | 42.8% |
| LMArena Search | — | 1199 |
| METR Time Horizons | 36.1% | — |
| Vending-Bench 2 | — | 1,473 |
Reasoning GPT-5.1 leads
GPT-4: 17.8 (#289), GPT-5.1: 39.8 (#58)
| Benchmark | GPT-4 | GPT-5.1 |
|---|---|---|
| Chess Puzzles | 4% | 32% |
| LMArena Hard Prompts | 1241 | 1457 |
| Mystery Game Puzzles | 12% | 19% |
| DTBench | 62.7% | 90.1% |
| LMCA | 17.1% | 43.9% |
| Epoch Capabilities Index | 125.89 | 149.64 |
| ForecastBench | 57.8 | 58.1 |
| ARC-AGI-2 | — | 17.6% |
| SimpleBench | — | 53.2% |
| ARC-AGI-1 | — | 72.8% |
| CritPt | — | 4.9% |
| EnigmaEval | — | 11.2% |
| LiveBench Reasoning | — | 95.8% |
| LiveBench Data Analysis | — | 72.1% |
| BIG-Bench Hard | 75.1% | — |
| HellaSwag | 95.3% | — |
| LiveBench | — | 78.8% |
| WinoGrande | 87.5% | — |
Math GPT-5.1 leads
GPT-4: 10.8 (#309), GPT-5.1: 52.2 (#51)
| Benchmark | GPT-4 | GPT-5.1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.1% | 88.6% |
| LMArena Math | 1269 | 1447 |
| Omni-MATH | — | 46.4% |
| LiveBench Math | — | 94.5% |
| MATH Level 5 | 23% | — |
| FrontierMath (Feb 2025 set) | — | 31% |
| FrontierMath Tier 4 (v1) | — | 12.5% |
| GSM8K | 92% | — |
Knowledge GPT-5.1 leads
GPT-4: 18.4 (#282), GPT-5.1: 50.6 (#71)
| Benchmark | GPT-4 | GPT-5.1 |
|---|---|---|
| GPQA Diamond | 35.7% | 87.6% |
| LMArena Expert | 1211 | 1470 |
| Humanity's Last Exam | — | 23.7% |
| SimpleQA Verified | — | 48% |
| MMLU-Pro | — | 57.9% |
| Vectara Hallucination Rate | — | 10.9% |
| GPQA (HELM) | — | 44.2% |
| MMLU | 86.4% | — |
| TriviaQA | 84.8% | — |
Multimodal Not comparable
GPT-4: —, GPT-5.1: 44.8 (#19)
| Benchmark | GPT-4 | GPT-5.1 |
|---|---|---|
| LMArena Vision | — | 1250 |
| VPCT | — | 58.7% |
| LMArena Document | — | 1403 |
Multilingual GPT-5.1 leads
GPT-4: 40.6 (#215), GPT-5.1: 53.8 (#56)
| Benchmark | GPT-4 | GPT-5.1 |
|---|---|---|
| LMArena Non-English | 1246 | 1431 |
| LMArena Chinese | 1242 | 1495 |
| LMArena French | 1283 | 1450 |
| LMArena German | 1251 | 1438 |
| LMArena Japanese | 1209 | 1453 |
| LMArena Korean | 1184 | 1401 |
| LMArena Russian | 1251 | 1435 |
| LMArena Spanish | 1261 | 1433 |
Instruction Following GPT-5.1 leads
GPT-4: 65.3 (#222), GPT-5.1: 83.9 (#1)
| Benchmark | GPT-4 | GPT-5.1 |
|---|---|---|
| LMArena Instruction Following | 1241 | 1443 |
| LiveBench Instruction Following | — | 93.3% |
| IFEval | — | 93.5% |
Long Context GPT-5.1 leads
GPT-4: 37.7 (#212), GPT-5.1: 47.6 (#14)
| Benchmark | GPT-4 | GPT-5.1 |
|---|---|---|
| LMArena Longer Query | 1244 | 1447 |
| CL-bench | — | 23.7% |
| CL-bench Life | — | 17.3% |
Writing & Preference GPT-5.1 leads
GPT-4: 34.9 (#268), GPT-5.1: 64.5 (#55)
| Benchmark | GPT-4 | GPT-5.1 |
|---|---|---|
| LMArena Text | 1263 | 1443 |
| LMArena Creative Writing | 1244 | 1427 |
| LMArena Multi-Turn | 1257 | 1450 |
| EQ-Bench Creative Writing | 752 | — |
| WildBench | — | 86.3% |
| LiveBench Language | — | 80.2% |
Frequently asked questions
Is GPT-4 better than GPT-5.1?
GPT-5.1 is the stronger model overall, scoring 49.0 to 29.1 on the Noometry Index.
Which is cheaper, GPT-4 or GPT-5.1?
GPT-5.1 is cheaper. It lists at $1.25 per million input tokens and $10 per million output tokens; GPT-4 lists at $30 and $60.
Is GPT-4 or GPT-5.1 better for coding?
GPT-5.1 scores higher on coding benchmarks: 46.4 versus 31.6 in the Noometry coding category.
Which has the bigger context window?
GPT-5.1 does, with 400K tokens against 8K.
How many benchmarks do GPT-4 and GPT-5.1 share?
26 benchmarks have published results for both models. GPT-4 has 38 scored results on Noometry and GPT-5.1 has 63.