Model comparison
Gemini 3 Pro vs Llama 4 Maverick
Gemini 3 Pro is the stronger model overall, scoring 54.8 to 30.9 on the Noometry Index.
Last verified . 43 shared benchmarks.
Google
MetaSummary
- They share 43 benchmarks with published results for both. Gemini 3 Pro scores higher in 10 categories and Llama 4 Maverick in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3 Pro leads 52.5 to 10.1.
- The biggest single-benchmark swing is NYT Connections (extended): 94.4% for Gemini 3 Pro and 8% for Llama 4 Maverick.
- Llama 4 Maverick has downloadable open weights; the other is API-only.
Side by side
| Gemini 3 Pro | Llama 4 Maverick | |
|---|---|---|
| Provider | Meta | |
| Noometry Index | 54.8 | 30.9 |
| Released | 2025-11-18 | 2025-04-05 |
| Weights | Proprietary | Open |
| Context window | — | 128K |
| Max output | — | 4K |
| Input $ / M tokens | — | $0.19 |
| Output $ / M tokens | — | $0.65 |
| Results tracked | 67 | 54 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Gemini 3 Pro leads
Gemini 3 Pro: 51.6 (#39), Llama 4 Maverick: 26.6 (#324)
| Benchmark | Gemini 3 Pro | Llama 4 Maverick |
|---|---|---|
| SWE-bench Verified (bash only) | 74.2% | 21% |
| WeirdML | 69.9% | 24.5% |
| LMArena Coding | 1481 | 1302 |
| ALE-Bench | 1,177 | 172.97 |
| SWE-bench Verified | 72.9% | — |
| Aider Polyglot | — | 15.6% |
| LMArena WebDev | 1440 | — |
| SWE-bench Multilingual | 68.7% | — |
| SciCode | — | 33.1% |
| GSO | 18.6% | — |
| BigCodeBench Instruct | — | 49.7% |
| BigCodeBench Complete | — | 61.4% |
| AlgoTune | 1.83 | — |
Agentic & Tool Use Gemini 3 Pro leads
Gemini 3 Pro: 40.6 (#23), Llama 4 Maverick: 28.2 (#91)
| Benchmark | Gemini 3 Pro | Llama 4 Maverick |
|---|---|---|
| Berkeley Function Calling Leaderboard | 72.5% | 37.3% |
| Terminal-Bench | 69.4% | — |
| GDPval | 40.3% | — |
| Remote Labor Index | 1.3% | — |
| τ²-bench Airline | 80.5% | — |
| τ²-bench Banking | 18% | — |
| τ²-bench Retail | 75.9% | — |
| τ²-bench Telecom | 91% | — |
| DeepResearch Bench | 46.3% | — |
| BALROG | 58.1% | — |
| LMArena Search | 1207 | — |
| METR Time Horizons | 71% | — |
| Vending-Bench 2 | 5,478 | — |
Reasoning Gemini 3 Pro leads
Gemini 3 Pro: 52.5 (#31), Llama 4 Maverick: 10.1 (#342)
| Benchmark | Gemini 3 Pro | Llama 4 Maverick |
|---|---|---|
| ARC-AGI-2 | 31.1% | 0% |
| SimpleBench | 76.4% | 27.7% |
| Kagi LLM Benchmark | 80.1% | 55.9% |
| NYT Connections (extended) | 94.4% | 8% |
| ARC-AGI-1 | 75% | 4.4% |
| CritPt | 6.9% | 0% |
| EnigmaEval | 18.2% | 0.6% |
| LMArena Hard Prompts | 1480 | 1281 |
| Epoch Capabilities Index | 152.92 | 132.2 |
| ForecastBench | 61.2 | 57.5 |
| Chess Puzzles | 31% | — |
| DTBench | — | 61.9% |
| LMCA | — | 15.9% |
Math Gemini 3 Pro leads
Gemini 3 Pro: 49.9 (#59), Llama 4 Maverick: 26.0 (#262)
| Benchmark | Gemini 3 Pro | Llama 4 Maverick |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.4% | 20.6% |
| Omni-MATH | 55.5% | 42.2% |
| LMArena Math | 1476 | 1299 |
| FrontierMath (Feb 2025 set) | 37.6% | 0.7% |
| MathArena Final-Answer Competitions | 67% | — |
| ProofBench | 20% | — |
| MATH Level 5 | — | 73% |
| FrontierMath Tier 4 (v1) | 18.8% | — |
Knowledge Gemini 3 Pro leads
Gemini 3 Pro: 64.4 (#16), Llama 4 Maverick: 33.4 (#204)
| Benchmark | Gemini 3 Pro | Llama 4 Maverick |
|---|---|---|
| GPQA Diamond | 92.6% | 67% |
| Humanity's Last Exam | 37.5% | 5.7% |
| MMLU-Pro | 90.3% | 81% |
| Vectara Hallucination Rate | 13.6% | 8.2% |
| GPQA (HELM) | 80.3% | 65% |
| LMArena Expert | 1475 | 1259 |
| Confabulations | — | 22.6% |
Multimodal Gemini 3 Pro leads
Gemini 3 Pro: 57.6 (#2), Llama 4 Maverick: 31.6 (#105)
| Benchmark | Gemini 3 Pro | Llama 4 Maverick |
|---|---|---|
| LMArena Vision | 1305 | 1142 |
| GeoBench | 84% | 52% |
| VPCT | 91% | — |
| LMArena Document | 1434 | — |
| SpatialViz-Bench | — | 31.8% |
Multilingual Gemini 3 Pro leads
Gemini 3 Pro: 56.9 (#16), Llama 4 Maverick: 42.2 (#195)
| Benchmark | Gemini 3 Pro | Llama 4 Maverick |
|---|---|---|
| LMArena Non-English | 1474 | 1269 |
| LMArena Chinese | 1523 | 1277 |
| LMArena French | 1492 | 1259 |
| LMArena German | 1515 | 1291 |
| LMArena Japanese | 1510 | 1207 |
| LMArena Korean | 1448 | 1203 |
| LMArena Russian | 1493 | 1286 |
| LMArena Spanish | 1470 | 1293 |
Instruction Following Gemini 3 Pro leads
Gemini 3 Pro: 76.3 (#45), Llama 4 Maverick: 71.7 (#146)
| Benchmark | Gemini 3 Pro | Llama 4 Maverick |
|---|---|---|
| IFEval | 87.7% | 90.8% |
| LMArena Instruction Following | 1458 | 1267 |
Long Context Gemini 3 Pro leads
Gemini 3 Pro: 44.0 (#79), Llama 4 Maverick: 31.4 (#279)
| Benchmark | Gemini 3 Pro | Llama 4 Maverick |
|---|---|---|
| LMArena Longer Query | 1471 | 1280 |
| Fiction.LiveBench | — | 46.2% |
| CL-bench | 15.8% | — |
Writing & Preference Gemini 3 Pro leads
Gemini 3 Pro: 66.4 (#35), Llama 4 Maverick: 38.8 (#252)
| Benchmark | Gemini 3 Pro | Llama 4 Maverick |
|---|---|---|
| LMArena Text | 1479 | 1287 |
| LMArena Creative Writing | 1482 | 1267 |
| EQ-Bench Creative Writing | 1525 | 860 |
| WildBench | 85.9% | 80% |
| LMArena Multi-Turn | 1484 | 1289 |
| Short-Story Creative Writing | — | 62% |
Frequently asked questions
Is Gemini 3 Pro better than Llama 4 Maverick?
Gemini 3 Pro is the stronger model overall, scoring 54.8 to 30.9 on the Noometry Index.
Is Gemini 3 Pro or Llama 4 Maverick better for coding?
Gemini 3 Pro scores higher on coding benchmarks: 51.6 versus 26.6 in the Noometry coding category.
How many benchmarks do Gemini 3 Pro and Llama 4 Maverick share?
43 benchmarks have published results for both models. Gemini 3 Pro has 67 scored results on Noometry and Llama 4 Maverick has 54.