Model comparison
Gemini 2.0 Flash (Feb 2025) vs GPT-5.4
GPT-5.4 is the stronger model overall, scoring 59.4 to 35.1 on the Noometry Index.
Last verified . 29 shared benchmarks.
GoogleSummary
- They share 29 benchmarks with published results for both. Gemini 2.0 Flash (Feb 2025) scores higher in 0 categories and GPT-5.4 in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GPT-5.4 leads 61.8 to 15.2.
- The biggest single-benchmark swing is ARC-AGI-2: 1.3% for Gemini 2.0 Flash (Feb 2025) and 74% for GPT-5.4.
Side by side
| Gemini 2.0 Flash (Feb 2025) | GPT-5.4 | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 35.1 | 59.4 |
| Released | 2024-12-06 | 2026-03-05 |
| Weights | Proprietary | Proprietary |
| Context window | — | 1.05M |
| Max output | — | 128K |
| Input $ / M tokens | — | $2.50 |
| Output $ / M tokens | — | $15 |
| Results tracked | 54 | 68 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GPT-5.4 leads
Gemini 2.0 Flash (Feb 2025): 28.4 (#315), GPT-5.4: 52.6 (#33)
| Benchmark | Gemini 2.0 Flash (Feb 2025) | GPT-5.4 |
|---|---|---|
| WeirdML | 25.8% | 77.7% |
| LMArena Coding | 1350 | 1497 |
| SWE-bench Verified | — | 76.9% |
| DeepSWE | — | 51.8% |
| SWE-bench Verified (bash only) | 13.5% | — |
| Aider Polyglot | 38.2% | — |
| LMArena WebDev | — | 1465 |
| SciCode | — | 56.6% |
| GSO | — | 31.4% |
| BigCodeBench Instruct | 45.9% | — |
| LiveBench Coding | 63.4% | — |
| MirrorCode | — | 15.6% |
| BigCodeBench Complete | 59.9% | — |
| CadEval | 30% | — |
| ALE-Bench | — | 1,607 |
| AlgoTune | — | 1.85 |
Agentic & Tool Use GPT-5.4 leads
Gemini 2.0 Flash (Feb 2025): 28.1 (#92), GPT-5.4: 46.5 (#13)
| Benchmark | Gemini 2.0 Flash (Feb 2025) | GPT-5.4 |
|---|---|---|
| Terminal-Bench | — | 81.8% |
| APEX-Agents | — | 52.4% |
| TheAgentCompany | 11.4% | — |
| τ²-bench Banking | — | 39.4% |
| DeepResearch Bench | — | 35.1% |
| PostTrainBench | — | 19% |
| GBAEval | — | 45.1% |
| LMArena Search | — | 1197 |
| METR Time Horizons | — | 74.3% |
| Vending-Bench 2 | — | 6,144 |
Reasoning GPT-5.4 leads
Gemini 2.0 Flash (Feb 2025): 15.2 (#318), GPT-5.4: 61.8 (#19)
| Benchmark | Gemini 2.0 Flash (Feb 2025) | GPT-5.4 |
|---|---|---|
| ARC-AGI-2 | 1.3% | 74% |
| Kagi LLM Benchmark | 37.8% | 63.8% |
| EnigmaEval | 1.1% | 16% |
| LMArena Hard Prompts | 1346 | 1485 |
| DTBench | 63.2% | 94.4% |
| Epoch Capabilities Index | 135.36 | 156.81 |
| SimpleBench | 31.1% | — |
| NYT Connections (extended) | — | 91.3% |
| ARC-AGI-1 | — | 93.7% |
| CritPt | — | 23.4% |
| Chess Puzzles | — | 44% |
| Thematic Generalization | — | 80% |
| EBR-Bench | — | 25.4% |
| LiveBench Reasoning | 78.2% | — |
| Mystery Game Puzzles | — | 37% |
| LiveBench Data Analysis | 69.4% | — |
| LMCA | — | 52% |
| ForecastBench | — | 59.5 |
| LiveBench | 66.9% | — |
Math GPT-5.4 leads
Gemini 2.0 Flash (Feb 2025): 37.9 (#146), GPT-5.4: 73.5 (#19)
| Benchmark | Gemini 2.0 Flash (Feb 2025) | GPT-5.4 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 57.8% | 97.8% |
| LMArena Math | 1352 | 1488 |
| FrontierMath (Feb 2025 set) | 1.7% | 47.6% |
| FrontierMath (Tiers 1-3) | — | 78.6% |
| FrontierMath Tier 4 | — | 49% |
| MathArena Final-Answer Competitions | — | 83.1% |
| ProofBench | — | 56% |
| Omni-MATH | 45.9% | — |
| LiveBench Math | 75.8% | — |
| MATH Level 5 | 82.2% | — |
| FrontierMath Tier 4 (v1) | — | 27.1% |
Knowledge GPT-5.4 leads
Gemini 2.0 Flash (Feb 2025): 32.0 (#213), GPT-5.4: 65.3 (#14)
| Benchmark | Gemini 2.0 Flash (Feb 2025) | GPT-5.4 |
|---|---|---|
| GPQA Diamond | 64.1% | 93.3% |
| Humanity's Last Exam | 6.6% | 36.2% |
| LMArena Expert | 1339 | 1507 |
| SimpleQA Verified | — | 45.1% |
| MMLU-Pro | 73.7% | — |
| Confabulations | 12.4% | — |
| Vectara Hallucination Rate | — | 7% |
| GPQA (HELM) | 55.6% | — |
| MMLU | 79.7% | — |
Multimodal GPT-5.4 leads
Gemini 2.0 Flash (Feb 2025): 36.5 (#79), GPT-5.4: 43.7 (#20)
| Benchmark | Gemini 2.0 Flash (Feb 2025) | GPT-5.4 |
|---|---|---|
| LMArena Vision | 1158 | 1303 |
| GeoBench | 77% | — |
| Blueprint-Bench 2 | — | 27.1% |
| Furniture Assembly | — | 37.5% |
| LMArena Document | — | 1471 |
Multilingual GPT-5.4 leads
Gemini 2.0 Flash (Feb 2025): 47.4 (#149), GPT-5.4: 56.2 (#23)
| Benchmark | Gemini 2.0 Flash (Feb 2025) | GPT-5.4 |
|---|---|---|
| LMArena Non-English | 1342 | 1465 |
| LMArena Chinese | 1373 | 1519 |
| LMArena French | 1391 | 1493 |
| LMArena German | 1353 | 1472 |
| LMArena Japanese | 1294 | 1485 |
| LMArena Korean | 1313 | 1448 |
| LMArena Russian | 1351 | 1480 |
| LMArena Spanish | 1363 | 1454 |
Instruction Following GPT-5.4 leads
Gemini 2.0 Flash (Feb 2025): 74.4 (#97), GPT-5.4: 77.1 (#27)
| Benchmark | Gemini 2.0 Flash (Feb 2025) | GPT-5.4 |
|---|---|---|
| LMArena Instruction Following | 1336 | 1469 |
| LiveBench Instruction Following | 85.8% | — |
| IFEval | 84.1% | — |
Long Context GPT-5.4 leads
Gemini 2.0 Flash (Feb 2025): 38.1 (#203), GPT-5.4: 50.3 (#8)
| Benchmark | Gemini 2.0 Flash (Feb 2025) | GPT-5.4 |
|---|---|---|
| LMArena Longer Query | 1344 | 1473 |
| Fiction.LiveBench | 61.1% | — |
| CL-bench | — | 27.9% |
| CL-bench Life | — | 21.7% |
Writing & Preference GPT-5.4 leads
Gemini 2.0 Flash (Feb 2025): 49.5 (#190), GPT-5.4: 71.9 (#17)
| Benchmark | Gemini 2.0 Flash (Feb 2025) | GPT-5.4 |
|---|---|---|
| LMArena Text | 1354 | 1469 |
| LMArena Creative Writing | 1340 | 1439 |
| EQ-Bench Creative Writing | 1128 | 1840 |
| LMArena Multi-Turn | 1350 | 1482 |
| Short-Story Creative Writing | 73.8% | — |
| WildBench | 80% | — |
| EQ-Bench 4 | — | 1272 |
| LiveBench Language | 51.3% | — |
Frequently asked questions
Is Gemini 2.0 Flash (Feb 2025) better than GPT-5.4?
GPT-5.4 is the stronger model overall, scoring 59.4 to 35.1 on the Noometry Index.
Is Gemini 2.0 Flash (Feb 2025) or GPT-5.4 better for coding?
GPT-5.4 scores higher on coding benchmarks: 52.6 versus 28.4 in the Noometry coding category.
How many benchmarks do Gemini 2.0 Flash (Feb 2025) and GPT-5.4 share?
29 benchmarks have published results for both models. Gemini 2.0 Flash (Feb 2025) has 54 scored results on Noometry and GPT-5.4 has 68.