Model comparison
Amazon Nova Lite vs DeepSeek V4 Flash
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 31.9 on the Noometry Index. Amazon Nova Lite costs 2.5× less per token, which makes it the better buy when DeepSeek V4 Flash's lead doesn't matter for your workload.
Last verified . 18 shared benchmarks.
DeepSeekSummary
- They share 18 benchmarks with published results for both. Amazon Nova Lite scores higher in 0 categories and DeepSeek V4 Flash in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Flash leads 53.7 to 19.3.
- Amazon Nova Lite is cheaper at $0.06 / $0.24 per million input/output tokens, against $0.15 / $0.60 for DeepSeek V4 Flash.
- DeepSeek V4 Flash accepts more context: 1M tokens versus 300K.
- DeepSeek V4 Flash has downloadable open weights; the other is API-only.
Side by side
| Amazon Nova Lite | DeepSeek V4 Flash | |
|---|---|---|
| Provider | Amazon | DeepSeek |
| Noometry Index | 31.9 | 53.6 |
| Released | 2024-12-03 | 2026-04-24 |
| Weights | Proprietary | Open |
| Context window | 300K | 1M |
| Max output | 10K | 393K |
| Input $ / M tokens | $0.06 | $0.15 |
| Output $ / M tokens | $0.24 | $0.60 |
| Results tracked | 34 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Flash leads
Amazon Nova Lite: 32.5 (#270), DeepSeek V4 Flash: 47.9 (#59)
| Benchmark | Amazon Nova Lite | DeepSeek V4 Flash |
|---|---|---|
| LMArena Coding | 1239 | 1457 |
| ALE-Bench | 236.25 | 1,306 |
| FrontierCode | — | 18.8% |
| LMArena WebDev | — | 1582 |
| SciCode | — | 49.9% |
| WeirdML | — | 63% |
| LiveBench Coding | 27.5% | — |
Reasoning DeepSeek V4 Flash leads
Amazon Nova Lite: 19.3 (#260), DeepSeek V4 Flash: 53.7 (#30)
| Benchmark | Amazon Nova Lite | DeepSeek V4 Flash |
|---|---|---|
| LMArena Hard Prompts | 1220 | 1444 |
| ARC-AGI-2 | — | 61.4% |
| SimpleBench | — | 61.1% |
| Kagi LLM Benchmark | — | 52.2% |
| NYT Connections (extended) | — | 89.6% |
| ARC-AGI-1 | — | 89% |
| CritPt | — | 16.6% |
| Chess Puzzles | — | 33% |
| LiveBench Reasoning | 36.7% | — |
| Mystery Game Puzzles | — | 34% |
| DTBench | — | 90.9% |
| LiveBench Data Analysis | 37.2% | — |
| LMCA | — | 41.7% |
| Epoch Capabilities Index | — | 154.49 |
| LiveBench | 36.4% | — |
Math DeepSeek V4 Flash leads
Amazon Nova Lite: 27.8 (#247), DeepSeek V4 Flash: 60.3 (#37)
| Benchmark | Amazon Nova Lite | DeepSeek V4 Flash |
|---|---|---|
| LMArena Math | 1227 | 1427 |
| FrontierMath (Tiers 1-3) | — | 57.5% |
| FrontierMath Tier 4 | — | 24.4% |
| MathArena Final-Answer Competitions | — | 76.5% |
| OTIS Mock AIME 2024-2025 | — | 94.4% |
| ProofBench | — | 56% |
| Omni-MATH | 23.3% | — |
| LiveBench Math | 36.7% | — |
Knowledge DeepSeek V4 Flash leads
Amazon Nova Lite: 26.3 (#259), DeepSeek V4 Flash: 55.4 (#48)
| Benchmark | Amazon Nova Lite | DeepSeek V4 Flash |
|---|---|---|
| LMArena Expert | 1201 | 1441 |
| GPQA Diamond | — | 91% |
| Humanity's Last Exam | 3.6% | — |
| SimpleQA Verified | — | 33.6% |
| MMLU-Pro | 60% | — |
| Vectara Hallucination Rate | 6.1% | — |
| GPQA (HELM) | 39.7% | — |
| MMLU | 77% | — |
Multimodal Not comparable
Amazon Nova Lite: 25.5 (#123), DeepSeek V4 Flash: —
| Benchmark | Amazon Nova Lite | DeepSeek V4 Flash |
|---|---|---|
| LMArena Vision | 990 | — |
Multilingual DeepSeek V4 Flash leads
Amazon Nova Lite: 38.0 (#232), DeepSeek V4 Flash: 53.0 (#72)
| Benchmark | Amazon Nova Lite | DeepSeek V4 Flash |
|---|---|---|
| LMArena Non-English | 1208 | 1420 |
| LMArena Chinese | 1225 | 1468 |
| LMArena French | 1237 | 1439 |
| LMArena German | 1229 | 1418 |
| LMArena Japanese | 1153 | 1406 |
| LMArena Korean | 1153 | 1384 |
| LMArena Russian | 1216 | 1428 |
| LMArena Spanish | 1226 | 1436 |
Instruction Following DeepSeek V4 Flash leads
Amazon Nova Lite: 59.3 (#255), DeepSeek V4 Flash: 74.9 (#81)
| Benchmark | Amazon Nova Lite | DeepSeek V4 Flash |
|---|---|---|
| LMArena Instruction Following | 1205 | 1421 |
| LiveBench Instruction Following | 54.1% | — |
| IFEval | 77.6% | — |
Long Context DeepSeek V4 Flash leads
Amazon Nova Lite: 37.4 (#217), DeepSeek V4 Flash: 43.8 (#85)
| Benchmark | Amazon Nova Lite | DeepSeek V4 Flash |
|---|---|---|
| LMArena Longer Query | 1234 | 1434 |
Writing & Preference DeepSeek V4 Flash leads
Amazon Nova Lite: 42.1 (#237), DeepSeek V4 Flash: 63.8 (#61)
| Benchmark | Amazon Nova Lite | DeepSeek V4 Flash |
|---|---|---|
| LMArena Text | 1229 | 1432 |
| LMArena Creative Writing | 1197 | 1403 |
| LMArena Multi-Turn | 1199 | 1449 |
| EQ-Bench Creative Writing | — | 1559 |
| WildBench | 75% | — |
| LiveBench Language | 25.9% | — |
Frequently asked questions
Is Amazon Nova Lite better than DeepSeek V4 Flash?
DeepSeek V4 Flash is the stronger model overall, scoring 53.6 to 31.9 on the Noometry Index. Amazon Nova Lite costs 2.5× less per token, which makes it the better buy when DeepSeek V4 Flash's lead doesn't matter for your workload.
Which is cheaper, Amazon Nova Lite or DeepSeek V4 Flash?
Amazon Nova Lite is cheaper. It lists at $0.06 per million input tokens and $0.24 per million output tokens; DeepSeek V4 Flash lists at $0.15 and $0.60.
Is Amazon Nova Lite or DeepSeek V4 Flash better for coding?
DeepSeek V4 Flash scores higher on coding benchmarks: 47.9 versus 32.5 in the Noometry coding category.
Which has the bigger context window?
DeepSeek V4 Flash does, with 1M tokens against 300K.
How many benchmarks do Amazon Nova Lite and DeepSeek V4 Flash share?
18 benchmarks have published results for both models. Amazon Nova Lite has 34 scored results on Noometry and DeepSeek V4 Flash has 41.