Model comparison
Gemini 2.5 Pro vs GPT-4.5
Gemini 2.5 Pro is the stronger model overall, scoring 45.0 to 37.2 on the Noometry Index.
Last verified . 41 shared benchmarks.
GoogleSummary
- They share 41 benchmarks with published results for both. Gemini 2.5 Pro scores higher in 9 categories and GPT-4.5 in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Gemini 2.5 Pro leads 56.0 to 32.5.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 84.7% for Gemini 2.5 Pro and 37.8% for GPT-4.5.
Side by side
| Gemini 2.5 Pro | GPT-4.5 | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 45.0 | 37.2 |
| Released | 2025-03-25 | 2025-02-27 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | — |
| Max output | 66K | — |
| Input $ / M tokens | $1.25 | — |
| Output $ / M tokens | $10 | — |
| Results tracked | 78 | 42 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Gemini 2.5 Pro: 42.4 (#101), GPT-4.5: 42.2 (#109)
| Benchmark | Gemini 2.5 Pro | GPT-4.5 |
|---|---|---|
| Aider Polyglot | 83.1% | 44.9% |
| WeirdML | 54% | 39.4% |
| LiveBench Coding | 85.9% | 75.2% |
| LMArena Coding | 1452 | 1396 |
| SWE-bench Verified | 57.6% | — |
| SWE-bench Verified (bash only) | 53.6% | — |
| LMArena WebDev | 1227 | — |
| SciCode | 42.8% | — |
| GSO | 3.9% | — |
| CadEval | 64% | — |
| ALE-Bench | 785.52 | — |
| AlgoTune | 1.51 | — |
Agentic & Tool Use Gemini 2.5 Pro leads
Gemini 2.5 Pro: 29.2 (#88), GPT-4.5: 27.9 (#97)
| Benchmark | Gemini 2.5 Pro | GPT-4.5 |
|---|---|---|
| Terminal-Bench | 32.6% | — |
| GDPval | 23.3% | — |
| Remote Labor Index | 0.8% | — |
| TheAgentCompany | 30.3% | — |
| τ²-bench Banking | 13.7% | — |
| Cybench | — | 17.5% |
| DeepResearch Bench | 42.8% | — |
| BALROG | 43.3% | — |
| LMArena Search | 1142 | — |
| METR Time Horizons | 55.4% | — |
| Vending-Bench 2 | 573.64 | — |
Reasoning Gemini 2.5 Pro leads
Gemini 2.5 Pro: 28.8 (#99), GPT-4.5: 13.9 (#330)
| Benchmark | Gemini 2.5 Pro | GPT-4.5 |
|---|---|---|
| ARC-AGI-2 | 4.9% | 0.8% |
| SimpleBench | 62.4% | 34.5% |
| ARC-AGI-1 | 41% | 10.3% |
| EnigmaEval | 5.6% | 3.2% |
| LiveBench Reasoning | 89.8% | 71.1% |
| LMArena Hard Prompts | 1455 | 1403 |
| LiveBench Data Analysis | 79.9% | 64.3% |
| Epoch Capabilities Index | 145.32 | 136.74 |
| ForecastBench | 61.3 | 61.7 |
| LiveBench | 82.3% | 69% |
| Kagi LLM Benchmark | 70.3% | — |
| CritPt | 2% | — |
| Chess Puzzles | 20% | — |
| DTBench | 82.4% | — |
| LMCA | 34.8% | — |
Math Too close to call
Gemini 2.5 Pro: 32.5 (#213), GPT-4.5: 32.6 (#211)
| Benchmark | Gemini 2.5 Pro | GPT-4.5 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 84.7% | 37.8% |
| LiveBench Math | 90.2% | 69.3% |
| LMArena Math | 1450 | 1412 |
| MATH Level 5 | 95.9% | 78.6% |
| FrontierMath (Tiers 1-3) | 24.6% | — |
| FrontierMath Tier 4 | 0% | — |
| Omni-MATH | 41.6% | — |
| FrontierMath (Feb 2025 set) | 14.1% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Gemini 2.5 Pro leads
Gemini 2.5 Pro: 56.0 (#46), GPT-4.5: 32.5 (#211)
| Benchmark | Gemini 2.5 Pro | GPT-4.5 |
|---|---|---|
| GPQA Diamond | 85.3% | 68.7% |
| Humanity's Last Exam | 21.6% | 5.4% |
| Confabulations | 10.6% | 13.6% |
| LMArena Expert | 1452 | 1394 |
| MMLU-Pro | 86.3% | — |
| Vectara Hallucination Rate | 7% | — |
| GPQA (HELM) | 74.9% | — |
Multimodal Gemini 2.5 Pro leads
Gemini 2.5 Pro: 45.2 (#18), GPT-4.5: 37.6 (#71)
| Benchmark | Gemini 2.5 Pro | GPT-4.5 |
|---|---|---|
| LMArena Vision | 1263 | 1195 |
| VPCT | 48% | 45% |
| GeoBench | 86% | — |
| LMArena Document | 1421 | — |
| SpatialViz-Bench | 44.7% | — |
Multilingual Gemini 2.5 Pro leads
Gemini 2.5 Pro: 55.3 (#31), GPT-4.5: 52.5 (#83)
| Benchmark | Gemini 2.5 Pro | GPT-4.5 |
|---|---|---|
| LMArena Non-English | 1451 | 1413 |
| LMArena Chinese | 1507 | 1421 |
| LMArena French | 1472 | 1418 |
| LMArena German | 1487 | 1457 |
| LMArena Japanese | 1461 | 1416 |
| LMArena Korean | 1434 | 1392 |
| LMArena Russian | 1461 | 1419 |
| LMArena Spanish | 1473 | — |
Instruction Following Gemini 2.5 Pro leads
Gemini 2.5 Pro: 75.0 (#75), GPT-4.5: 72.6 (#134)
| Benchmark | Gemini 2.5 Pro | GPT-4.5 |
|---|---|---|
| LiveBench Instruction Following | 80.6% | 72.3% |
| LMArena Instruction Following | 1437 | 1404 |
| IFEval | 84% | — |
Long Context Gemini 2.5 Pro leads
Gemini 2.5 Pro: 59.8 (#5), GPT-4.5: 40.4 (#155)
| Benchmark | Gemini 2.5 Pro | GPT-4.5 |
|---|---|---|
| Fiction.LiveBench | 91.7% | 63.9% |
| LMArena Longer Query | 1449 | 1406 |
Writing & Preference Gemini 2.5 Pro leads
Gemini 2.5 Pro: 63.7 (#62), GPT-4.5: 56.9 (#134)
| Benchmark | Gemini 2.5 Pro | GPT-4.5 |
|---|---|---|
| LMArena Text | 1458 | 1417 |
| LMArena Creative Writing | 1454 | 1394 |
| Short-Story Creative Writing | 83.8% | 75.6% |
| EQ-Bench Creative Writing | 1421 | 1258 |
| LMArena Multi-Turn | 1453 | 1444 |
| LiveBench Language | 67.8% | 61.5% |
| WildBench | 85.7% | — |
Frequently asked questions
Is Gemini 2.5 Pro better than GPT-4.5?
Gemini 2.5 Pro is the stronger model overall, scoring 45.0 to 37.2 on the Noometry Index.
Is Gemini 2.5 Pro or GPT-4.5 better for coding?
They score almost the same on coding (42.4 vs 42.2); test both on your own repository before choosing.
How many benchmarks do Gemini 2.5 Pro and GPT-4.5 share?
41 benchmarks have published results for both models. Gemini 2.5 Pro has 78 scored results on Noometry and GPT-4.5 has 42.