Model comparison
Gemini 2.5 Flash vs GPT-4o
Gemini 2.5 Flash is the stronger model overall, scoring 39.3 to 28.6 on the Noometry Index.
Last verified . 46 shared benchmarks.
GoogleSummary
- They share 46 benchmarks with published results for both. Gemini 2.5 Flash scores higher in 10 categories and GPT-4o in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where Gemini 2.5 Flash leads 39.9 to 10.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 73.1% for Gemini 2.5 Flash and 6.4% for GPT-4o.
- Gemini 2.5 Flash is cheaper at $0.30 / $2.50 per million input/output tokens, against $2.50 / $10 for GPT-4o.
- Gemini 2.5 Flash accepts more context: 1.05M tokens versus 128K.
Side by side
| Gemini 2.5 Flash | GPT-4o | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 39.3 | 28.6 |
| Released | 2025-04-17 | 2024-05-13 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 128K |
| Max output | 66K | 16K |
| Input $ / M tokens | $0.30 | $2.50 |
| Output $ / M tokens | $2.50 | $10 |
| Results tracked | 54 | 72 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 2.5 Flash leads
Gemini 2.5 Flash: 35.8 (#220), GPT-4o: 24.8 (#328)
| Benchmark | Gemini 2.5 Flash | GPT-4o |
|---|---|---|
| SWE-bench Verified (bash only) | 28.7% | 21.6% |
| Aider Polyglot | 55.1% | 45.3% |
| WeirdML | 41.9% | 25.1% |
| LMArena Coding | 1424 | 1297 |
| SWE-bench Verified | — | 31% |
| GSO | — | 0% |
| BigCodeBench Instruct | — | 51.1% |
| LiveBench Coding | — | 51.4% |
| BigCodeBench Complete | — | 61.1% |
| CadEval | — | 26% |
| ALE-Bench | 661.88 | — |
| HumanEval+ | — | 87.2% |
| MBPP+ | — | 72.2% |
Agentic & Tool Use Gemini 2.5 Flash leads
Gemini 2.5 Flash: 30.8 (#74), GPT-4o: 21.0 (#141)
| Benchmark | Gemini 2.5 Flash | GPT-4o |
|---|---|---|
| TheAgentCompany | 41.1% | 8.6% |
| BALROG | 33.5% | 32.3% |
| Terminal-Bench | 17.1% | — |
| Berkeley Function Calling Leaderboard | 56.2% | — |
| GDPval | — | 9.9% |
| Cybench | — | 12.5% |
| LMArena Search | — | 1006 |
| METR Time Horizons | — | 40.8% |
| Vending-Bench 2 | 548.84 | — |
Reasoning Gemini 2.5 Flash leads
Gemini 2.5 Flash: 18.1 (#286), GPT-4o: 9.4 (#343)
| Benchmark | Gemini 2.5 Flash | GPT-4o |
|---|---|---|
| ARC-AGI-2 | 2.5% | 0% |
| SimpleBench | 41.2% | 17.8% |
| ARC-AGI-1 | 33.3% | 4.5% |
| CritPt | 1.1% | 0% |
| EnigmaEval | 2.7% | 0.8% |
| LMArena Hard Prompts | 1422 | 1281 |
| DTBench | 76.5% | 64.5% |
| LMCA | 27.5% | 16.6% |
| Epoch Capabilities Index | 143.03 | 128.97 |
| ForecastBench | 60.6 | 57.7 |
| Kagi LLM Benchmark | 56.8% | — |
| Chess Puzzles | — | 13% |
| LiveBench Reasoning | — | 55.8% |
| LiveBench Data Analysis | — | 60.9% |
| LiveBench | — | 55.3% |
Math Gemini 2.5 Flash leads
Gemini 2.5 Flash: 39.9 (#98), GPT-4o: 10.6 (#312)
| Benchmark | Gemini 2.5 Flash | GPT-4o |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 73.1% | 6.4% |
| Omni-MATH | 38.5% | 29.3% |
| LMArena Math | 1415 | 1285 |
| FrontierMath (Feb 2025 set) | 4.8% | 0.3% |
| FrontierMath (Tiers 1-3) | — | 0.4% |
| LiveBench Math | — | 49.5% |
| MATH Level 5 | — | 53.3% |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge Gemini 2.5 Flash leads
Gemini 2.5 Flash: 36.4 (#168), GPT-4o: 28.8 (#242)
| Benchmark | Gemini 2.5 Flash | GPT-4o |
|---|---|---|
| Humanity's Last Exam | 12.1% | 2.7% |
| MMLU-Pro | 63.9% | 71.3% |
| Confabulations | 16.8% | 15.3% |
| Vectara Hallucination Rate | 7.8% | 9.6% |
| GPQA (HELM) | 39% | 52% |
| LMArena Expert | 1426 | 1250 |
| GPQA Diamond | — | 49.2% |
| SimpleQA Verified | — | 26% |
| MMLU | — | 88.1% |
Multimodal Gemini 2.5 Flash leads
Gemini 2.5 Flash: 41.8 (#32), GPT-4o: 34.5 (#91)
| Benchmark | Gemini 2.5 Flash | GPT-4o |
|---|---|---|
| LMArena Vision | 1253 | 1137 |
| GeoBench | 76% | 71% |
| VPCT | 46.2% | 40% |
| Video-MME | — | 71.9% |
| ScienceQA | — | 88.5% |
| SpatialViz-Bench | 36.9% | — |
Multilingual Gemini 2.5 Flash leads
Gemini 2.5 Flash: 52.3 (#88), GPT-4o: 43.2 (#186)
| Benchmark | Gemini 2.5 Flash | GPT-4o |
|---|---|---|
| LMArena Non-English | 1409 | 1283 |
| LMArena Chinese | 1450 | 1277 |
| LMArena French | 1433 | 1304 |
| LMArena German | 1418 | 1282 |
| LMArena Japanese | 1405 | 1257 |
| LMArena Korean | 1385 | 1234 |
| LMArena Russian | 1415 | 1286 |
| LMArena Spanish | 1421 | 1292 |
Instruction Following Gemini 2.5 Flash leads
Gemini 2.5 Flash: 75.7 (#54), GPT-4o: 66.6 (#207)
| Benchmark | Gemini 2.5 Flash | GPT-4o |
|---|---|---|
| IFEval | 89.8% | 81.7% |
| LMArena Instruction Following | 1405 | 1278 |
| LiveBench Instruction Following | — | 68.6% |
Long Context Gemini 2.5 Flash leads
Gemini 2.5 Flash: 47.5 (#17), GPT-4o: 39.4 (#179)
| Benchmark | Gemini 2.5 Flash | GPT-4o |
|---|---|---|
| Fiction.LiveBench | 77.8% | 66.7% |
| LMArena Longer Query | 1419 | 1289 |
Writing & Preference Gemini 2.5 Flash leads
Gemini 2.5 Flash: 53.8 (#157), GPT-4o: 52.6 (#166)
| Benchmark | Gemini 2.5 Flash | GPT-4o |
|---|---|---|
| LMArena Text | 1417 | 1300 |
| LMArena Creative Writing | 1400 | 1292 |
| Short-Story Creative Writing | 76.5% | 81.8% |
| WildBench | 81.7% | 82.8% |
| LMArena Multi-Turn | 1408 | 1302 |
| EQ-Bench Creative Writing | 1137 | — |
| LiveBench Language | — | 47.6% |
Frequently asked questions
Is Gemini 2.5 Flash better than GPT-4o?
Gemini 2.5 Flash is the stronger model overall, scoring 39.3 to 28.6 on the Noometry Index.
Which is cheaper, Gemini 2.5 Flash or GPT-4o?
Gemini 2.5 Flash is cheaper. It lists at $0.30 per million input tokens and $2.50 per million output tokens; GPT-4o lists at $2.50 and $10.
Is Gemini 2.5 Flash or GPT-4o better for coding?
Gemini 2.5 Flash scores higher on coding benchmarks: 35.8 versus 24.8 in the Noometry coding category.
Which has the bigger context window?
Gemini 2.5 Flash does, with 1.05M tokens against 128K.
How many benchmarks do Gemini 2.5 Flash and GPT-4o share?
46 benchmarks have published results for both models. Gemini 2.5 Flash has 54 scored results on Noometry and GPT-4o has 72.