Model comparison
Gemini 3 Flash Preview vs GPT-5.4 nano
Gemini 3 Flash Preview is the stronger model overall, scoring 52.3 to 41.9 on the Noometry Index. GPT-5.4 nano costs 2.4× less per token, which makes it the better buy when Gemini 3 Flash Preview's lead doesn't matter for your workload.
Last verified . 37 shared benchmarks.
GoogleSummary
- They share 37 benchmarks with published results for both. Gemini 3 Flash Preview scores higher in 9 categories and GPT-5.4 nano in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3 Flash Preview leads 49.2 to 23.8.
- The biggest single-benchmark swing is SimpleQA Verified: 66.8% for Gemini 3 Flash Preview and 11.7% for GPT-5.4 nano.
- GPT-5.4 nano is cheaper at $0.20 / $1.25 per million input/output tokens, against $0.50 / $3 for Gemini 3 Flash Preview.
- Gemini 3 Flash Preview accepts more context: 1.05M tokens versus 400K.
Side by side
| Gemini 3 Flash Preview | GPT-5.4 nano | |
|---|---|---|
| Provider | OpenAI | |
| Noometry Index | 52.3 | 41.9 |
| Released | 2025-12-17 | 2026-03-17 |
| Weights | Proprietary | Proprietary |
| Context window | 1.05M | 400K |
| Max output | 66K | 128K |
| Input $ / M tokens | $0.50 | $0.20 |
| Output $ / M tokens | $3 | $1.25 |
| Results tracked | 59 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3 Flash Preview leads
Gemini 3 Flash Preview: 51.5 (#35), GPT-5.4 nano: 43.1 (#82)
| Benchmark | Gemini 3 Flash Preview | GPT-5.4 nano |
|---|---|---|
| WeirdML | 61.6% | 49.2% |
| LMArena Coding | 1460 | 1404 |
| ALE-Bench | 1,367 | 1,005 |
| SWE-bench Verified | 75.4% | — |
| SWE-bench Verified (bash only) | 75.8% | — |
| LMArena WebDev | 1439 | — |
| SWE-bench Multilingual | 72.7% | — |
| SciCode | — | 46.9% |
| GSO | 9.8% | — |
Agentic & Tool Use Not comparable
Gemini 3 Flash Preview: 38.6 (#27), GPT-5.4 nano: —
| Benchmark | Gemini 3 Flash Preview | GPT-5.4 nano |
|---|---|---|
| Terminal-Bench | 64.3% | — |
| τ²-bench Airline | 82.5% | — |
| τ²-bench Banking | 27.3% | — |
| τ²-bench Retail | 76.8% | — |
| τ²-bench Telecom | 91.2% | — |
| DeepResearch Bench | 49.8% | — |
| BALROG | 48.1% | — |
| GDP.pdf | 10% | — |
| LMArena Search | 1198 | — |
| Vending-Bench 2 | 3,635 | — |
Reasoning Gemini 3 Flash Preview leads
Gemini 3 Flash Preview: 49.2 (#35), GPT-5.4 nano: 23.8 (#159)
| Benchmark | Gemini 3 Flash Preview | GPT-5.4 nano |
|---|---|---|
| ARC-AGI-2 | 33.6% | 5.7% |
| ARC-AGI-1 | 84.7% | 51.5% |
| Chess Puzzles | 40% | 30% |
| LMArena Hard Prompts | 1465 | 1380 |
| Mystery Game Puzzles | 26% | 9% |
| DTBench | 89.1% | 80.3% |
| LMCA | 43.1% | 36.9% |
| Epoch Capabilities Index | 151.83 | 145.78 |
| ForecastBench | 58.5 | 57.3 |
| SimpleBench | 61.1% | — |
| Kagi LLM Benchmark | — | 39.7% |
| NYT Connections (extended) | 83.1% | — |
| CritPt | — | 9.3% |
Math Gemini 3 Flash Preview leads
Gemini 3 Flash Preview: 51.4 (#42), GPT-5.4 nano: 40.7 (#76)
| Benchmark | Gemini 3 Flash Preview | GPT-5.4 nano |
|---|---|---|
| FrontierMath (Tiers 1-3) | 51.2% | 44.9% |
| FrontierMath Tier 4 | 17.1% | 12.2% |
| OTIS Mock AIME 2024-2025 | 95.6% | 87.8% |
| ProofBench | 15% | 5% |
| LMArena Math | 1472 | 1406 |
| FrontierMath (Feb 2025 set) | 35.6% | 25.9% |
| FrontierMath Tier 4 (v1) | 4.2% | 6.3% |
| MathArena Final-Answer Competitions | 67.6% | — |
Knowledge Gemini 3 Flash Preview leads
Gemini 3 Flash Preview: 58.8 (#30), GPT-5.4 nano: 41.8 (#94)
| Benchmark | Gemini 3 Flash Preview | GPT-5.4 nano |
|---|---|---|
| GPQA Diamond | 89.4% | 78.5% |
| SimpleQA Verified | 66.8% | 11.7% |
| Vectara Hallucination Rate | 13.5% | 3.1% |
| LMArena Expert | 1461 | 1397 |
Multimodal Gemini 3 Flash Preview leads
Gemini 3 Flash Preview: 45.3 (#15), GPT-5.4 nano: 36.0 (#73)
| Benchmark | Gemini 3 Flash Preview | GPT-5.4 nano |
|---|---|---|
| LMArena Vision | 1285 | 1196 |
| GeoBench | 88% | — |
| VPCT | 72.6% | — |
| Blueprint-Bench 2 | 0% | — |
| LMArena Document | 1413 | — |
Multilingual Gemini 3 Flash Preview leads
Gemini 3 Flash Preview: 55.9 (#26), GPT-5.4 nano: 48.8 (#136)
| Benchmark | Gemini 3 Flash Preview | GPT-5.4 nano |
|---|---|---|
| LMArena Non-English | 1458 | 1359 |
| LMArena Chinese | 1510 | 1391 |
| LMArena French | 1478 | 1396 |
| LMArena German | 1497 | 1368 |
| LMArena Japanese | 1488 | 1342 |
| LMArena Korean | 1443 | 1321 |
| LMArena Russian | 1480 | 1363 |
| LMArena Spanish | 1470 | 1368 |
Instruction Following Gemini 3 Flash Preview leads
Gemini 3 Flash Preview: 75.8 (#52), GPT-5.4 nano: 72.1 (#139)
| Benchmark | Gemini 3 Flash Preview | GPT-5.4 nano |
|---|---|---|
| LMArena Instruction Following | 1437 | 1362 |
Long Context Gemini 3 Flash Preview leads
Gemini 3 Flash Preview: 44.4 (#63), GPT-5.4 nano: 41.6 (#131)
| Benchmark | Gemini 3 Flash Preview | GPT-5.4 nano |
|---|---|---|
| LMArena Longer Query | 1452 | 1365 |
Writing & Preference Gemini 3 Flash Preview leads
Gemini 3 Flash Preview: 65.6 (#44), GPT-5.4 nano: 55.9 (#137)
| Benchmark | Gemini 3 Flash Preview | GPT-5.4 nano |
|---|---|---|
| LMArena Text | 1466 | 1372 |
| LMArena Creative Writing | 1457 | 1313 |
| LMArena Multi-Turn | 1471 | 1382 |
Frequently asked questions
Is Gemini 3 Flash Preview better than GPT-5.4 nano?
Gemini 3 Flash Preview is the stronger model overall, scoring 52.3 to 41.9 on the Noometry Index. GPT-5.4 nano costs 2.4× less per token, which makes it the better buy when Gemini 3 Flash Preview's lead doesn't matter for your workload.
Which is cheaper, Gemini 3 Flash Preview or GPT-5.4 nano?
GPT-5.4 nano is cheaper. It lists at $0.20 per million input tokens and $1.25 per million output tokens; Gemini 3 Flash Preview lists at $0.50 and $3.
Is Gemini 3 Flash Preview or GPT-5.4 nano better for coding?
Gemini 3 Flash Preview scores higher on coding benchmarks: 51.5 versus 43.1 in the Noometry coding category.
Which has the bigger context window?
Gemini 3 Flash Preview does, with 1.05M tokens against 400K.
How many benchmarks do Gemini 3 Flash Preview and GPT-5.4 nano share?
37 benchmarks have published results for both models. Gemini 3 Flash Preview has 59 scored results on Noometry and GPT-5.4 nano has 40.