Model comparison
GPT-5.2 vs o3-mini
GPT-5.2 is the stronger model overall, scoring 54.1 to 36.7 on the Noometry Index. o3-mini costs 2.5× less per token, which makes it the better buy when GPT-5.2's lead doesn't matter for your workload.
Last verified . 35 shared benchmarks.
Summary
- They share 35 benchmarks with published results for both. GPT-5.2 scores higher in 8 categories and o3-mini in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.2 leads 50.2 to 16.3.
- The biggest single-benchmark swing is ARC-AGI-1: 86.2% for GPT-5.2 and 34.5% for o3-mini.
- o3-mini is cheaper at $1.10 / $4.40 per million input/output tokens, against $1.75 / $14 for GPT-5.2.
- GPT-5.2 accepts more context: 400K tokens versus 200K.
Side by side
| GPT-5.2 | o3-mini | |
|---|---|---|
| Provider | OpenAI | OpenAI |
| Noometry Index | 54.1 | 36.7 |
| Released | 2025-12-11 | 2024-12-20 |
| Weights | Proprietary | Proprietary |
| Context window | 400K | 200K |
| Max output | 128K | 100K |
| Input $ / M tokens | $1.75 | $1.10 |
| Output $ / M tokens | $14 | $4.40 |
| Results tracked | 67 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.2 leads
GPT-5.2: 51.6 (#37), o3-mini: 40.8 (#132)
| Benchmark | GPT-5.2 | o3-mini |
|---|---|---|
| GSO | 27.4% | 1.3% |
| WeirdML | 72.2% | 43.7% |
| LMArena Coding | 1447 | 1378 |
| SWE-bench Verified | 73.8% | — |
| SWE-bench Verified (bash only) | 72.8% | — |
| Aider Polyglot | — | 60.4% |
| LMArena WebDev | 1416 | — |
| SWE-bench Multilingual | 66.7% | — |
| SciCode | — | 39.8% |
| LiveBench Coding | — | 82.7% |
| CadEval | — | 54% |
| ALE-Bench | 1,294 | — |
| AlgoTune | 2.05 | — |
Agentic & Tool Use GPT-5.2 leads
GPT-5.2: 40.2 (#24), o3-mini: 29.6 (#84)
| Benchmark | GPT-5.2 | o3-mini |
|---|---|---|
| Terminal-Bench | 64.9% | — |
| Berkeley Function Calling Leaderboard | 55.9% | — |
| GDPval | 49.7% | — |
| Remote Labor Index | 2.5% | — |
| τ²-bench Airline | 83% | — |
| τ²-bench Banking | 32.2% | — |
| τ²-bench Retail | 81.6% | — |
| τ²-bench Telecom | 89.7% | — |
| Cybench | — | 22.5% |
| DeepResearch Bench | 41.1% | — |
| LMArena Search | 1207 | — |
| METR Time Horizons | 75.3% | — |
| Vending-Bench 2 | 3,591 | — |
Reasoning GPT-5.2 leads
GPT-5.2: 50.2 (#35), o3-mini: 16.3 (#305)
| Benchmark | GPT-5.2 | o3-mini |
|---|---|---|
| ARC-AGI-2 | 52.9% | 3% |
| SimpleBench | 45.8% | 22.8% |
| ARC-AGI-1 | 86.2% | 34.5% |
| Chess Puzzles | 49% | 17% |
| LMArena Hard Prompts | 1445 | 1366 |
| Mystery Game Puzzles | 23% | 7% |
| DTBench | 90.9% | 68.8% |
| LMCA | 43.9% | 19% |
| Epoch Capabilities Index | 153.45 | 140.34 |
| ForecastBench | 60.1 | 59.6 |
| Kagi LLM Benchmark | 73.3% | — |
| NYT Connections (extended) | 83.6% | — |
| CritPt | — | 0.3% |
| EnigmaEval | 10.4% | — |
| EBR-Bench | 23% | — |
| LiveBench Reasoning | — | 89.6% |
| LiveBench Data Analysis | — | 70.6% |
| LiveBench | — | 75.9% |
Math GPT-5.2 leads
GPT-5.2: 60.0 (#38), o3-mini: 28.1 (#244)
| Benchmark | GPT-5.2 | o3-mini |
|---|---|---|
| FrontierMath (Tiers 1-3) | 67.4% | 18.6% |
| FrontierMath Tier 4 | 31.7% | 0% |
| OTIS Mock AIME 2024-2025 | 96.1% | 76.9% |
| LMArena Math | 1440 | 1396 |
| FrontierMath (Feb 2025 set) | 40.7% | 12.4% |
| FrontierMath Tier 4 (v1) | 18.8% | 4.2% |
| MathArena Final-Answer Competitions | 72% | — |
| ProofBench | 15% | — |
| LiveBench Math | — | 77.3% |
| MATH Level 5 | — | 96.5% |
Knowledge GPT-5.2 leads
GPT-5.2: 59.3 (#32), o3-mini: 38.3 (#146)
| Benchmark | GPT-5.2 | o3-mini |
|---|---|---|
| GPQA Diamond | 91.4% | 77% |
| SimpleQA Verified | 37.1% | 15.3% |
| LMArena Expert | 1445 | 1364 |
| Humanity's Last Exam | 27.8% | — |
| Confabulations | — | 17.9% |
| Vectara Hallucination Rate | 8.4% | — |
Multimodal Not comparable
GPT-5.2: 51.3 (#7), o3-mini: —
| Benchmark | GPT-5.2 | o3-mini |
|---|---|---|
| LMArena Vision | 1268 | — |
| VPCT | 84% | — |
| Furniture Assembly | 38.3% | — |
| LMArena Document | 1405 | — |
Multilingual GPT-5.2 leads
GPT-5.2: 53.4 (#67), o3-mini: 45.7 (#164)
| Benchmark | GPT-5.2 | o3-mini |
|---|---|---|
| LMArena Non-English | 1425 | 1319 |
| LMArena Chinese | 1460 | 1379 |
| LMArena French | 1455 | 1334 |
| LMArena German | 1448 | 1303 |
| LMArena Japanese | 1420 | 1286 |
| LMArena Korean | 1392 | 1314 |
| LMArena Russian | 1440 | 1304 |
| LMArena Spanish | 1433 | 1321 |
Instruction Following Too close to call
GPT-5.2: 74.7 (#89), o3-mini: 75.1 (#72)
| Benchmark | GPT-5.2 | o3-mini |
|---|---|---|
| LMArena Instruction Following | 1417 | 1337 |
| LiveBench Instruction Following | — | 84.4% |
Long Context GPT-5.2 leads
GPT-5.2: 44.0 (#78), o3-mini: 33.8 (#256)
| Benchmark | GPT-5.2 | o3-mini |
|---|---|---|
| LMArena Longer Query | 1428 | 1343 |
| Fiction.LiveBench | — | 50% |
| CL-bench | 18.2% | — |
Writing & Preference GPT-5.2 leads
GPT-5.2: 66.8 (#32), o3-mini: 50.3 (#182)
| Benchmark | GPT-5.2 | o3-mini |
|---|---|---|
| LMArena Text | 1439 | 1337 |
| LMArena Creative Writing | 1401 | 1286 |
| LMArena Multi-Turn | 1458 | 1320 |
| Short-Story Creative Writing | — | 61.7% |
| EQ-Bench Creative Writing | 1703 | — |
| LiveBench Language | — | 50.7% |
Frequently asked questions
Is GPT-5.2 better than o3-mini?
GPT-5.2 is the stronger model overall, scoring 54.1 to 36.7 on the Noometry Index. o3-mini costs 2.5× less per token, which makes it the better buy when GPT-5.2's lead doesn't matter for your workload.
Which is cheaper, GPT-5.2 or o3-mini?
o3-mini is cheaper. It lists at $1.10 per million input tokens and $4.40 per million output tokens; GPT-5.2 lists at $1.75 and $14.
Is GPT-5.2 or o3-mini better for coding?
GPT-5.2 scores higher on coding benchmarks: 51.6 versus 40.8 in the Noometry coding category.
Which has the bigger context window?
GPT-5.2 does, with 400K tokens against 200K.
How many benchmarks do GPT-5.2 and o3-mini share?
35 benchmarks have published results for both models. GPT-5.2 has 67 scored results on Noometry and o3-mini has 51.