Model comparison
Amazon Nova Pro vs DeepSeek V4 Pro
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 31.0 on the Noometry Index.
Last verified . 19 shared benchmarks.
DeepSeekSummary
- They share 19 benchmarks with published results for both. Amazon Nova Pro scores higher in 0 categories and DeepSeek V4 Pro in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 20.0.
- DeepSeek V4 Pro is cheaper at $0.66 / $1.98 per million input/output tokens, against $0.80 / $3.20 for Amazon Nova Pro.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 300K.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| Amazon Nova Pro | DeepSeek V4 Pro | |
|---|---|---|
| Provider | Amazon | DeepSeek |
| Noometry Index | 31.0 | 54.3 |
| Released | 2024-12-03 | 2026-04-24 |
| Weights | Proprietary | Open |
| Context window | 300K | 1M |
| Max output | 10K | 393K |
| Input $ / M tokens | $0.80 | $0.66 |
| Output $ / M tokens | $3.20 | $1.98 |
| Results tracked | 38 | 48 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
Amazon Nova Pro: 35.1 (#229), DeepSeek V4 Pro: 52.4 (#34)
| Benchmark | Amazon Nova Pro | DeepSeek V4 Pro |
|---|---|---|
| LMArena Coding | 1270 | 1470 |
| SWE-bench Verified | — | 77.6% |
| FrontierCode | — | 28.6% |
| LMArena WebDev | — | 1582 |
| SciCode | — | 51% |
| WeirdML | — | 66.2% |
| LiveBench Coding | 38.1% | — |
| ALE-Bench | — | 1,403 |
Agentic & Tool Use DeepSeek V4 Pro leads
Amazon Nova Pro: 16.7 (#147), DeepSeek V4 Pro: 32.8 (#58)
| Benchmark | Amazon Nova Pro | DeepSeek V4 Pro |
|---|---|---|
| APEX-Agents | — | 47.3% |
| Berkeley Function Calling Leaderboard | 25% | — |
| TheAgentCompany | 1.7% | — |
| Vending-Bench 2 | — | 3,285 |
Reasoning DeepSeek V4 Pro leads
Amazon Nova Pro: 20.0 (#243), DeepSeek V4 Pro: 56.5 (#24)
| Benchmark | Amazon Nova Pro | DeepSeek V4 Pro |
|---|---|---|
| LMArena Hard Prompts | 1246 | 1461 |
| Epoch Capabilities Index | 123.8 | 155.31 |
| ARC-AGI-2 | — | 61.3% |
| Kagi LLM Benchmark | — | 53.5% |
| NYT Connections (extended) | — | 91.3% |
| ARC-AGI-1 | — | 90.5% |
| CritPt | — | 18% |
| Chess Puzzles | — | 47% |
| LiveBench Reasoning | 32.6% | — |
| Mystery Game Puzzles | — | 43% |
| DTBench | — | 93.9% |
| LiveBench Data Analysis | 48.3% | — |
| LMCA | — | 45.5% |
| Surface Evolver Bench | — | 40% |
| ForecastBench | — | 56.1 |
| LiveBench | 43.5% | — |
Math DeepSeek V4 Pro leads
Amazon Nova Pro: 28.5 (#243), DeepSeek V4 Pro: 64.8 (#30)
| Benchmark | Amazon Nova Pro | DeepSeek V4 Pro |
|---|---|---|
| LMArena Math | 1252 | 1455 |
| FrontierMath (Tiers 1-3) | — | 64.6% |
| FrontierMath Tier 4 | — | 26.8% |
| MathArena Final-Answer Competitions | — | 76.6% |
| OTIS Mock AIME 2024-2025 | — | 98.6% |
| ProofBench | — | 50% |
| Omni-MATH | 24.2% | — |
| LiveBench Math | 38% | — |
Knowledge DeepSeek V4 Pro leads
Amazon Nova Pro: 27.4 (#250), DeepSeek V4 Pro: 59.5 (#31)
| Benchmark | Amazon Nova Pro | DeepSeek V4 Pro |
|---|---|---|
| Vectara Hallucination Rate | 5.1% | 8.6% |
| LMArena Expert | 1211 | 1464 |
| GPQA Diamond | — | 91.7% |
| Humanity's Last Exam | 4.4% | — |
| SimpleQA Verified | — | 52.9% |
| MMLU-Pro | 67.3% | — |
| Confabulations | 30.1% | — |
| GPQA (HELM) | 44.6% | — |
| MMLU | 82% | — |
Multimodal Not comparable
Amazon Nova Pro: 25.0 (#126), DeepSeek V4 Pro: —
| Benchmark | Amazon Nova Pro | DeepSeek V4 Pro |
|---|---|---|
| LMArena Vision | 980 | — |
Multilingual DeepSeek V4 Pro leads
Amazon Nova Pro: 39.7 (#223), DeepSeek V4 Pro: 54.4 (#45)
| Benchmark | Amazon Nova Pro | DeepSeek V4 Pro |
|---|---|---|
| LMArena Non-English | 1234 | 1439 |
| LMArena Chinese | 1244 | 1486 |
| LMArena French | 1271 | 1472 |
| LMArena German | 1243 | 1458 |
| LMArena Japanese | 1200 | 1445 |
| LMArena Korean | 1203 | 1447 |
| LMArena Russian | 1240 | 1453 |
| LMArena Spanish | 1182 | 1458 |
Instruction Following DeepSeek V4 Pro leads
Amazon Nova Pro: 64.9 (#226), DeepSeek V4 Pro: 76.1 (#47)
| Benchmark | Amazon Nova Pro | DeepSeek V4 Pro |
|---|---|---|
| LMArena Instruction Following | 1235 | 1448 |
| LiveBench Instruction Following | 67.1% | — |
| IFEval | 81.5% | — |
Long Context DeepSeek V4 Pro leads
Amazon Nova Pro: 38.1 (#205), DeepSeek V4 Pro: 45.0 (#51)
| Benchmark | Amazon Nova Pro | DeepSeek V4 Pro |
|---|---|---|
| LMArena Longer Query | 1255 | 1458 |
| CL-bench Life | — | 13.5% |
Writing & Preference DeepSeek V4 Pro leads
Amazon Nova Pro: 43.9 (#226), DeepSeek V4 Pro: 65.5 (#46)
| Benchmark | Amazon Nova Pro | DeepSeek V4 Pro |
|---|---|---|
| LMArena Text | 1259 | 1451 |
| LMArena Creative Writing | 1212 | 1446 |
| LMArena Multi-Turn | 1246 | 1467 |
| Short-Story Creative Writing | 60.5% | — |
| EQ-Bench Creative Writing | — | 1553 |
| WildBench | 77.7% | — |
| EQ-Bench 4 | — | 1166 |
| LiveBench Language | 37% | — |
Frequently asked questions
Is Amazon Nova Pro better than DeepSeek V4 Pro?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 31.0 on the Noometry Index.
Which is cheaper, Amazon Nova Pro or DeepSeek V4 Pro?
DeepSeek V4 Pro is cheaper. It lists at $0.66 per million input tokens and $1.98 per million output tokens; Amazon Nova Pro lists at $0.80 and $3.20.
Is Amazon Nova Pro or DeepSeek V4 Pro better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 35.1 in the Noometry coding category.
Which has the bigger context window?
DeepSeek V4 Pro does, with 1M tokens against 300K.
How many benchmarks do Amazon Nova Pro and DeepSeek V4 Pro share?
19 benchmarks have published results for both models. Amazon Nova Pro has 38 scored results on Noometry and DeepSeek V4 Pro has 48.