Model comparison
Amazon Nova Pro vs Gemma 3 12B
Gemma 3 12B is the stronger model overall, scoring 32.1 to 31.0 on the Noometry Index.
Last verified . 15 shared benchmarks.
GoogleSummary
- They share 15 benchmarks with published results for both. Amazon Nova Pro scores higher in 4 categories and Gemma 3 12B in 5 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Gemma 3 12B leads 25.5 to 16.7.
- The biggest single-benchmark swing is Berkeley Function Calling Leaderboard: 25% for Amazon Nova Pro and 30.4% for Gemma 3 12B.
- Gemma 3 12B is cheaper at $0.05 / $0.15 per million input/output tokens, against $0.80 / $3.20 for Amazon Nova Pro.
- Amazon Nova Pro accepts more context: 300K tokens versus 131K.
- Gemma 3 12B has downloadable open weights; the other is API-only.
Side by side
| Amazon Nova Pro | Gemma 3 12B | |
|---|---|---|
| Provider | Amazon | |
| Noometry Index | 31.0 | 32.1 |
| Released | 2024-12-03 | 2025-03-12 |
| Weights | Proprietary | Open |
| Context window | 300K | 131K |
| Max output | 10K | 8K |
| Input $ / M tokens | $0.80 | $0.05 |
| Output $ / M tokens | $3.20 | $0.15 |
| Results tracked | 38 | 24 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Amazon Nova Pro leads
Amazon Nova Pro: 35.1 (#229), Gemma 3 12B: 31.7 (#280)
| Benchmark | Amazon Nova Pro | Gemma 3 12B |
|---|---|---|
| LMArena Coding | 1270 | 1281 |
| SciCode | — | 17.4% |
| LiveBench Coding | 38.1% | — |
Agentic & Tool Use Gemma 3 12B leads
Amazon Nova Pro: 16.7 (#147), Gemma 3 12B: 25.5 (#108)
| Benchmark | Amazon Nova Pro | Gemma 3 12B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 25% | 30.4% |
| TheAgentCompany | 1.7% | — |
Reasoning Amazon Nova Pro leads
Amazon Nova Pro: 20.0 (#243), Gemma 3 12B: 15.7 (#313)
| Benchmark | Amazon Nova Pro | Gemma 3 12B |
|---|---|---|
| LMArena Hard Prompts | 1246 | 1309 |
| Epoch Capabilities Index | 123.8 | 123.5 |
| CritPt | — | 0% |
| Chess Puzzles | — | 0% |
| LiveBench Reasoning | 32.6% | — |
| DTBench | — | 48.8% |
| LiveBench Data Analysis | 48.3% | — |
| LMCA | — | 4.5% |
| LiveBench | 43.5% | — |
Math Amazon Nova Pro leads
Amazon Nova Pro: 28.5 (#243), Gemma 3 12B: 22.3 (#279)
| Benchmark | Amazon Nova Pro | Gemma 3 12B |
|---|---|---|
| LMArena Math | 1252 | 1307 |
| OTIS Mock AIME 2024-2025 | — | 16.7% |
| Omni-MATH | 24.2% | — |
| LiveBench Math | 38% | — |
Knowledge Too close to call
Amazon Nova Pro: 27.4 (#250), Gemma 3 12B: 26.5 (#257)
| Benchmark | Amazon Nova Pro | Gemma 3 12B |
|---|---|---|
| Vectara Hallucination Rate | 5.1% | 4.4% |
| LMArena Expert | 1211 | 1248 |
| GPQA Diamond | — | 39.5% |
| Humanity's Last Exam | 4.4% | — |
| MMLU-Pro | 67.3% | — |
| Confabulations | 30.1% | — |
| GPQA (HELM) | 44.6% | — |
| MMLU | 82% | — |
Multimodal Not comparable
Amazon Nova Pro: 25.0 (#126), Gemma 3 12B: —
| Benchmark | Amazon Nova Pro | Gemma 3 12B |
|---|---|---|
| LMArena Vision | 980 | — |
| MindCube | — | 46.7% |
Multilingual Gemma 3 12B leads
Amazon Nova Pro: 39.7 (#223), Gemma 3 12B: 45.7 (#165)
| Benchmark | Amazon Nova Pro | Gemma 3 12B |
|---|---|---|
| LMArena Non-English | 1234 | 1318 |
| LMArena German | 1243 | 1370 |
| LMArena Russian | 1240 | 1335 |
| LMArena Chinese | 1244 | — |
| LMArena French | 1271 | — |
| LMArena Japanese | 1200 | — |
| LMArena Korean | 1203 | — |
| LMArena Spanish | 1182 | — |
Instruction Following Gemma 3 12B leads
Amazon Nova Pro: 64.9 (#226), Gemma 3 12B: 68.6 (#186)
| Benchmark | Amazon Nova Pro | Gemma 3 12B |
|---|---|---|
| LMArena Instruction Following | 1235 | 1299 |
| LiveBench Instruction Following | 67.1% | — |
| IFEval | 81.5% | — |
Long Context Gemma 3 12B leads
Amazon Nova Pro: 38.1 (#205), Gemma 3 12B: 40.0 (#162)
| Benchmark | Amazon Nova Pro | Gemma 3 12B |
|---|---|---|
| LMArena Longer Query | 1255 | 1317 |
Writing & Preference Gemma 3 12B leads
Amazon Nova Pro: 43.9 (#226), Gemma 3 12B: 47.5 (#209)
| Benchmark | Amazon Nova Pro | Gemma 3 12B |
|---|---|---|
| LMArena Text | 1259 | 1334 |
| LMArena Creative Writing | 1212 | 1331 |
| LMArena Multi-Turn | 1246 | 1334 |
| Short-Story Creative Writing | 60.5% | — |
| EQ-Bench Creative Writing | — | 1126 |
| WildBench | 77.7% | — |
| LiveBench Language | 37% | — |
Frequently asked questions
Is Amazon Nova Pro better than Gemma 3 12B?
Gemma 3 12B is the stronger model overall, scoring 32.1 to 31.0 on the Noometry Index.
Which is cheaper, Amazon Nova Pro or Gemma 3 12B?
Gemma 3 12B is cheaper. It lists at $0.05 per million input tokens and $0.15 per million output tokens; Amazon Nova Pro lists at $0.80 and $3.20.
Is Amazon Nova Pro or Gemma 3 12B better for coding?
Amazon Nova Pro scores higher on coding benchmarks: 35.1 versus 31.7 in the Noometry coding category.
Which has the bigger context window?
Amazon Nova Pro does, with 300K tokens against 131K.
How many benchmarks do Amazon Nova Pro and Gemma 3 12B share?
15 benchmarks have published results for both models. Amazon Nova Pro has 38 scored results on Noometry and Gemma 3 12B has 24.