Nova Pro 1.0 vs Gemma 3 4B
Updated October 21, 2026. 33 benchmarks have a result for both. Coding is ranked on the Vibe Coding Index; every other board is shown as published.
Decision reading
Neither model has a coding score yet, so the two cannot be ranked on coding.
Nova Pro 1.0 has the edge where the context window is larger, 300K against 131K tokens, and where what matters is Humanity's Last Exam (Artificial Analysis run) or LMArena Text: hard prompts, where it leads.
- Shared results
- 33
- Nova Pro 1.0 only
- 9
- Gemma 3 4B only
- 6
- Like-for-like categories
- 7 of 9
Shared results can support a head-to-head reading. Results only one model has describe coverage, not superiority.
Which one for your work
Each card is decided by one number both models carry. Where only one has it, the card says so and makes no call.
Coding
Vibe Coding Index, six public coding benchmarks
No call
Neither has a coding score yet.
One side not measured
Intelligence
AA Intelligence Index
Nova Pro 1.0
Nova Pro 1.0 leads, 7.0 against 4.8.
Same board, both measured
Web design
Design Arena: websites
No call
Only Nova Pro 1.0 has a Design Arena: websites result, 796.
Directional only
Long documents
Listed context window
Nova Pro 1.0
Nova Pro 1.0 takes 300K tokens against 131K.
Listed by OpenRouter
Everyday cost
1K input + 500 output tokens
Gemma 3 4B
Nova Pro 1.0 $0.0024, Gemma 3 4B $0.0001 per turn.
List prices
At a glance
- Coding index
- Nova Pro 1.0 not benchmarked, Gemma 3 4B not benchmarked.
- Widest category
- Knowledge, 14.1 points apart on average over 5 shared benchmarks.
- Widest single gap
- MMLU Pro (Artificial Analysis run): Nova Pro 1.0 69.1% against Gemma 3 4B 41.7%, 27.4 points apart on a 0 to 100 scale.
- Output price
- Nova Pro 1.0 $3.20 against Gemma 3 4B $0.075 per million output tokens, 42.7x apart.
- Context window
- Nova Pro 1.0 takes 300K tokens, Gemma 3 4B 131K.
Benchmark by benchmark
First the score: boards each model leads, per category, where both have a result. Then every board, a category to a fold; bars are each result on a 0 to 100 scale, so compare the two in a row, not one row with another.
| Category | Nova Pro 1.0 | Bars | Gemma 3 4B | Ties |
|---|---|---|---|---|
| Coding | 4 | 0 | ||
| Intelligence | 1 | 0 | ||
| Agentic work | 1 | 0 | ||
| Reasoning | 2 | 3 | ||
| Math | 2 | 2 | ||
| Knowledge | 1 | 4 | ||
| General | 2 | 11 | ||
| All shared boards | 13 | 20 |
CodingLike-for-likeSix of these boards feed the Vibe Coding Index40
| Benchmark | Nova Pro 1.0 | Gemma 3 4B | Leader |
|---|---|---|---|
| AA Coding Index | Not measured | 2.7 | Only Gemma 3 4B measured |
| LiveCodeBench (Artificial Analysis run) | 23.3% | 11.2% | Nova Pro 1.0 |
| Terminal-Bench Hard (Artificial Analysis run) | 6.1% | 0.8% | Nova Pro 1.0 |
| LMArena Text: coding | 1270 | 1230 | Nova Pro 1.0 |
| LMArena Text: software and IT | 1271 | 1269 | Nova Pro 1.0 |
| Neither model has a result on the other 33 boards here | |||
IntelligenceLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Nova Pro 1.0 | Gemma 3 4B | Leader |
|---|---|---|---|
| AA Intelligence Index | 7.0 | 4.8 | Nova Pro 1.0 |
| Epoch Capabilities Index | Not measured | 115.9 | Only Gemma 3 4B measured |
Agentic workLike-for-likePublic boards, shown as published, kept out of the index10
| Benchmark | Nova Pro 1.0 | Gemma 3 4B | Leader |
|---|---|---|---|
| τ²-Bench Telecom | 14.0% | 5.0% | Nova Pro 1.0 |
| Neither model has a result on the other 38 boards here | |||
ReasoningLike-for-likePublic boards, shown as published, kept out of the index23
| Benchmark | Nova Pro 1.0 | Gemma 3 4B | Leader |
|---|---|---|---|
| GPQA Diamond (Artificial Analysis run) | 49.9% | 29.1% | Nova Pro 1.0 |
| Humanity's Last Exam (Artificial Analysis run) | 3.2% | 5.3% | Gemma 3 4B |
| GPQA Diamond (Epoch AI run) | Not measured | 23.2% | Only Gemma 3 4B measured |
| LMArena Text: hard prompts | 1246 | 1253 | Gemma 3 4B |
| LMArena Text: expert | 1211 | 1223 | Gemma 3 4B |
| LMArena Text: hard prompts in English | 1259 | 1253 | Nova Pro 1.0 |
| Neither model has a result on the other 15 boards here | |||
MathLike-for-likePublic boards, shown as published, kept out of the index22
| Benchmark | Nova Pro 1.0 | Gemma 3 4B | Leader |
|---|---|---|---|
| AA Math Index | 7.0 | 12.7 | Gemma 3 4B |
| AIME 2025 (Artificial Analysis run) | 7.0% | 12.7% | Gemma 3 4B |
| OTIS Mock AIME | Not measured | 7.5% | Only Gemma 3 4B measured |
| LMArena Text: math | 1252 | 1239 | Nova Pro 1.0 |
| LMArena Text: mathematical work | 1261 | 1256 | Nova Pro 1.0 |
| Neither model has a result on the other 8 boards here | |||
KnowledgeLike-for-likePublic boards, shown as published, kept out of the index14
| Benchmark | Nova Pro 1.0 | Gemma 3 4B | Leader |
|---|---|---|---|
| MMLU Pro (Artificial Analysis run) | 69.1% | 41.7% | Nova Pro 1.0 |
| MMLU | 82.0% | Not measured | Only Nova Pro 1.0 measured |
| LMArena Text: science | 1255 | 1306 | Gemma 3 4B |
| LMArena Text: medicine and healthcare | 1260 | 1342 | Gemma 3 4B |
| LMArena Text: legal and government | 1281 | 1327 | Gemma 3 4B |
| LMArena Text: business and finance | 1229 | 1311 | Gemma 3 4B |
| Neither model has a result on the other 11 boards here | |||
Web designDirectional onlyPublic boards, shown as published, kept out of the index1 board, none shared
| Benchmark | Nova Pro 1.0 | Gemma 3 4B | Leader |
|---|---|---|---|
| Design Arena: websites | 796 | Not measured | Only Nova Pro 1.0 measured |
| Neither model has a result on the other 8 boards here | |||
GeneralLike-for-likePublic boards, shown as published, kept out of the index211
| Benchmark | Nova Pro 1.0 | Gemma 3 4B | Leader |
|---|---|---|---|
| IFBench | 38.1% | 28.3% | Nova Pro 1.0 |
| Long context reasoning (AA-LCR) | 21.0% | 6.7% | Nova Pro 1.0 |
| DTBench | Not measured | 50.9% | Only Gemma 3 4B measured |
| LMCA | Not measured | 2.8% | Only Gemma 3 4B measured |
| LMArena Text | 1259 | 1291 | Gemma 3 4B |
| LMArena Text: instruction following | 1235 | 1239 | Gemma 3 4B |
| LMArena Text: creative writing | 1212 | 1270 | Gemma 3 4B |
| LMArena Vision | 980 | Not measured | Only Nova Pro 1.0 measured |
| LMArena Text: English | 1276 | 1300 | Gemma 3 4B |
| LMArena Text: non-English | 1233 | 1273 | Gemma 3 4B |
| LMArena Text: Chinese | 1244 | Not measured | Only Nova Pro 1.0 measured |
| LMArena Text: German | 1243 | 1281 | Gemma 3 4B |
| LMArena Text: Japanese | 1200 | Not measured | Only Nova Pro 1.0 measured |
| LMArena Text: Korean | 1203 | Not measured | Only Nova Pro 1.0 measured |
| LMArena Text: French | 1271 | Not measured | Only Nova Pro 1.0 measured |
| LMArena Text: Spanish | 1182 | Not measured | Only Nova Pro 1.0 measured |
| LMArena Text: Russian | 1240 | 1294 | Gemma 3 4B |
| LMArena Text: multi-turn | 1246 | 1255 | Gemma 3 4B |
| LMArena Text: longer queries | 1255 | 1273 | Gemma 3 4B |
| LMArena Text: writing and language | 1247 | 1288 | Gemma 3 4B |
| LMArena Text: entertainment and media | 1210 | 1233 | Gemma 3 4B |
| LMArena Vision: English | 987 | Not measured | Only Nova Pro 1.0 measured |
| Neither model has a result on the other 24 boards here | |||
Image generationNot comparablePublic boards, shown as published, kept out of the index0 boards, none shared
| Benchmark | Nova Pro 1.0 | Gemma 3 4B | Leader |
|---|---|---|---|
| Neither model has a result on 5 boards here | |||
What each workload costs
Three fixed token mixes at each model's listed price, so per-token rates become a bill.
Chat turn
1K input + 500 output tokens
Gemma 3 4B costs less
Repository review
50K input + 3K output tokens
Gemma 3 4B costs less
Agent loop
200K input + 10K output tokens
Gemma 3 4B costs less
Listed OpenRouter rates, without cache or batch discounts. A workload that does not fit a model's context window would need splitting.
Specification differences
The leading value in a row is marked; a row with a missing value names no leader. Our two rows are marked in red.
| Specification | Amazon Nova Pro 1.0 | Google Gemma 3 4B |
|---|---|---|
| Maker | Amazon | |
| Weights | No public weights listed | Open weights |
| Released | December 5, 2024 | March 13, 2025 |
| Context window | 300K | 131K |
| Input $/Mtok | $0.80 | $0.045 |
| Output $/Mtok | $3.20 | $0.075 |
| Coding index | Not benchmarked | Not benchmarked |
| Independent measurements behind it | 0 | 0 |
| Cost per solved task | Not computable | Not computable |
| Source | OpenRouter listing | OpenRouter listing |
Questions
Is Nova Pro 1.0 better than Gemma 3 4B for coding?
Neither model has a coding score yet, so the two cannot be ranked on coding. Nova Pro 1.0 has the edge where the context window is larger, 300K against 131K tokens, and where what matters is Humanity's Last Exam (Artificial Analysis run) or LMArena Text: hard prompts, where it leads.
Which is cheaper, Nova Pro 1.0 or Gemma 3 4B?
Nova Pro 1.0 costs $0.80 per million input tokens and $3.20 per million output tokens; Gemma 3 4B costs $0.045 and $0.075. Cost per solved task is known for neither of them.
Which has the bigger context window, Nova Pro 1.0 or Gemma 3 4B?
Nova Pro 1.0, with 300K tokens against 131K.
How many benchmarks is the Nova Pro 1.0 vs Gemma 3 4B comparison based on?
33 benchmarks have a result for both models as of October 21, 2026. Nova Pro 1.0's coding index rests on 0 independent benchmarks and Gemma 3 4B's on 0. Where only one model has a result, the page says so instead of naming a winner.