Model comparison
Gemini 3.7 Flash vs GPT-5.4
Gemini 3.7 Flash and GPT-5.4 score almost the same on the Noometry Index (59.8 vs 59.4), so choose on price, context window or the category you care about most.
Last verified . 40 shared benchmarks.
GoogleSummary
- They share 40 benchmarks with published results for both. Gemini 3.7 Flash scores higher in 5 categories and GPT-5.4 in 5 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3.7 Flash leads 70.0 to 61.8.
- The biggest single-benchmark swing is SimpleQA Verified: 69.2% for Gemini 3.7 Flash and 45.1% for GPT-5.4.
- Gemini 3.7 Flash is cheaper at $0.75 / $3.75 per million input/output tokens, against $2.50 / $15 for GPT-5.4.
- GPT-5.4 accepts more context: 1.05M tokens versus 1.05M.
Side by side
| Gemini 3.7 Flash | GPT-5.4 | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 59.8 | 59.4 |
| Released | 2026-08-13 | 2026-03-05 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 1.05M |
| Max output | 66K | 128K |
| Input $ / M tokens | $0.75 | $2.50 |
| Output $ / M tokens | $3.75 | $15 |
| Results tracked | 44 | 68 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3.7 Flash leads
Gemini 3.7 Flash: 56.2 (#22), GPT-5.4: 52.6 (#33)
| Benchmark | Gemini 3.7 Flash | GPT-5.4 |
|---|---|---|
| DeepSWE | 65.5% | 51.8% |
| LMArena WebDev | 1592 | 1465 |
| SciCode | 59.8% | 56.6% |
| LMArena Coding | 1497 | 1497 |
| ALE-Bench | 904.3 | 1,607 |
| SWE-bench Verified | — | 76.9% |
| FrontierCode | 43.6% | — |
| FrontierSWE | 20.3% | — |
| GSO | — | 31.4% |
| WeirdML | — | 77.7% |
| MirrorCode | — | 15.6% |
| AlgoTune | — | 1.85 |
Agentic & Tool Use GPT-5.4 leads
Gemini 3.7 Flash: 42.1 (#19), GPT-5.4: 46.5 (#13)
| Benchmark | Gemini 3.7 Flash | GPT-5.4 |
|---|---|---|
| APEX-Agents | 67.8% | 52.4% |
| Terminal-Bench | — | 81.8% |
| Remote Labor Index | 5% | — |
| τ²-bench Banking | — | 39.4% |
| DeepResearch Bench | — | 35.1% |
| PostTrainBench | — | 19% |
| GBAEval | — | 45.1% |
| GDP.pdf | 23.8% | — |
| LMArena Search | — | 1197 |
| METR Time Horizons | — | 74.3% |
| Vending-Bench 2 | — | 6,144 |
Reasoning Gemini 3.7 Flash leads
Gemini 3.7 Flash: 70.0 (#15), GPT-5.4: 61.8 (#19)
| Benchmark | Gemini 3.7 Flash | GPT-5.4 |
|---|---|---|
| ARC-AGI-2 | 84.6% | 74% |
| NYT Connections (extended) | 94% | 91.3% |
| ARC-AGI-1 | 95.5% | 93.7% |
| CritPt | 14.3% | 23.4% |
| Chess Puzzles | 47% | 44% |
| LMArena Hard Prompts | 1494 | 1485 |
| Mystery Game Puzzles | 37% | 37% |
| DTBench | 96.8% | 94.4% |
| LMCA | 50.4% | 52% |
| Epoch Capabilities Index | 157.27 | 156.81 |
| Kagi LLM Benchmark | — | 63.8% |
| EnigmaEval | — | 16% |
| Thematic Generalization | — | 80% |
| EBR-Bench | — | 25.4% |
| ForecastBench | — | 59.5 |
Math GPT-5.4 leads
Gemini 3.7 Flash: 69.6 (#23), GPT-5.4: 73.5 (#19)
| Benchmark | Gemini 3.7 Flash | GPT-5.4 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 71.6% | 78.6% |
| FrontierMath Tier 4 | 36.6% | 49% |
| OTIS Mock AIME 2024-2025 | 97.2% | 97.8% |
| ProofBench | 58% | 56% |
| LMArena Math | 1507 | 1488 |
| MathArena Final-Answer Competitions | — | 83.1% |
| FrontierMath (Feb 2025 set) | — | 47.6% |
| FrontierMath Tier 4 (v1) | — | 27.1% |
Knowledge Gemini 3.7 Flash leads
Gemini 3.7 Flash: 69.7 (#5), GPT-5.4: 65.3 (#14)
| Benchmark | Gemini 3.7 Flash | GPT-5.4 |
|---|---|---|
| GPQA Diamond | 94.8% | 93.3% |
| SimpleQA Verified | 69.2% | 45.1% |
| LMArena Expert | 1508 | 1507 |
| Humanity's Last Exam | — | 36.2% |
| Vectara Hallucination Rate | — | 7% |
Multimodal GPT-5.4 leads
Gemini 3.7 Flash: 37.3 (#73), GPT-5.4: 43.7 (#20)
| Benchmark | Gemini 3.7 Flash | GPT-5.4 |
|---|---|---|
| LMArena Vision | 1316 | 1303 |
| Furniture Assembly | 26.7% | 37.5% |
| Blueprint-Bench 2 | — | 27.1% |
| LMArena Document | — | 1471 |
Multilingual Gemini 3.7 Flash leads
Gemini 3.7 Flash: 57.6 (#7), GPT-5.4: 56.2 (#23)
| Benchmark | Gemini 3.7 Flash | GPT-5.4 |
|---|---|---|
| LMArena Non-English | 1484 | 1465 |
| LMArena Chinese | 1548 | 1519 |
| LMArena French | 1505 | 1493 |
| LMArena German | 1498 | 1472 |
| LMArena Japanese | 1512 | 1485 |
| LMArena Korean | 1483 | 1448 |
| LMArena Russian | 1516 | 1480 |
| LMArena Spanish | 1503 | 1454 |
Instruction Following Too close to call
Gemini 3.7 Flash: 77.7 (#15), GPT-5.4: 77.1 (#27)
| Benchmark | Gemini 3.7 Flash | GPT-5.4 |
|---|---|---|
| LMArena Instruction Following | 1483 | 1469 |
Long Context GPT-5.4 leads
Gemini 3.7 Flash: 45.7 (#30), GPT-5.4: 50.3 (#8)
| Benchmark | Gemini 3.7 Flash | GPT-5.4 |
|---|---|---|
| LMArena Longer Query | 1492 | 1473 |
| CL-bench | — | 27.9% |
| CL-bench Life | — | 21.7% |
Writing & Preference Too close to call
Gemini 3.7 Flash: 71.2 (#20), GPT-5.4: 71.9 (#17)
| Benchmark | Gemini 3.7 Flash | GPT-5.4 |
|---|---|---|
| LMArena Text | 1486 | 1469 |
| LMArena Creative Writing | 1490 | 1439 |
| EQ-Bench Creative Writing | 1723 | 1840 |
| LMArena Multi-Turn | 1489 | 1482 |
| EQ-Bench 4 | — | 1272 |
Frequently asked questions
Is Gemini 3.7 Flash better than GPT-5.4?
Gemini 3.7 Flash and GPT-5.4 score almost the same on the Noometry Index (59.8 vs 59.4), so choose on price, context window or the category you care about most.
Which is cheaper, Gemini 3.7 Flash or GPT-5.4?
Gemini 3.7 Flash is cheaper. It lists at $0.75 per million input tokens and $3.75 per million output tokens; GPT-5.4 lists at $2.50 and $15.
Is Gemini 3.7 Flash or GPT-5.4 better for coding?
Gemini 3.7 Flash scores higher on coding benchmarks: 56.2 versus 52.6 in the Noometry coding category.
Which has the bigger context window?
GPT-5.4 does, with 1.05M tokens against 1.05M.
How many benchmarks do Gemini 3.7 Flash and GPT-5.4 share?
40 benchmarks have published results for both models. Gemini 3.7 Flash has 44 scored results on Noometry and GPT-5.4 has 68.