Model comparison
Amazon Nova Lite vs DeepSeek-V3.1-Terminus
DeepSeek-V3.1-Terminus is the stronger model overall, scoring 43.1 to 31.9 on the Noometry Index. Amazon Nova Lite costs 4.3× less per token, which makes it the better buy when DeepSeek-V3.1-Terminus's lead doesn't matter for your workload.
Last verified . 11 shared benchmarks.
DeepSeekSummary
- They share 11 benchmarks with published results for both. Amazon Nova Lite scores higher in 0 categories and DeepSeek-V3.1-Terminus in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where DeepSeek-V3.1-Terminus leads 61.0 to 42.1.
- Amazon Nova Lite is cheaper at $0.06 / $0.24 per million input/output tokens, against $0.27 / $1 for DeepSeek-V3.1-Terminus.
- Amazon Nova Lite accepts more context: 300K tokens versus 164K.
- DeepSeek-V3.1-Terminus has downloadable open weights; the other is API-only.
Side by side
| Amazon Nova Lite | DeepSeek-V3.1-Terminus | |
|---|---|---|
| Provider | Amazon | DeepSeek |
| Noometry Index | 31.9 | 43.1 |
| Released | 2024-12-03 | 2025-09-22 |
| Weights | Proprietary | Open |
| Context window | 300K | 164K |
| Max output | 10K | 147K |
| Input $ / M tokens | $0.06 | $0.27 |
| Output $ / M tokens | $0.24 | $1 |
| Results tracked | 34 | 16 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek-V3.1-Terminus leads
Amazon Nova Lite: 32.5 (#270), DeepSeek-V3.1-Terminus: 42.0 (#113)
| Benchmark | Amazon Nova Lite | DeepSeek-V3.1-Terminus |
|---|---|---|
| LMArena Coding | 1239 | 1426 |
| ALE-Bench | 236.25 | 745.17 |
| SciCode | — | 40.6% |
| LiveBench Coding | 27.5% | — |
Reasoning DeepSeek-V3.1-Terminus leads
Amazon Nova Lite: 19.3 (#260), DeepSeek-V3.1-Terminus: 26.4 (#133)
| Benchmark | Amazon Nova Lite | DeepSeek-V3.1-Terminus |
|---|---|---|
| LMArena Hard Prompts | 1220 | 1426 |
| Kagi LLM Benchmark | — | 57.4% |
| CritPt | — | 1.7% |
| LiveBench Reasoning | 36.7% | — |
| DTBench | — | 81.3% |
| LiveBench Data Analysis | 37.2% | — |
| LMCA | — | 28.6% |
| LiveBench | 36.4% | — |
Math DeepSeek-V3.1-Terminus leads
Amazon Nova Lite: 27.8 (#247), DeepSeek-V3.1-Terminus: 38.5 (#137)
| Benchmark | Amazon Nova Lite | DeepSeek-V3.1-Terminus |
|---|---|---|
| LMArena Math | 1227 | 1402 |
| Omni-MATH | 23.3% | — |
| LiveBench Math | 36.7% | — |
Knowledge Not comparable
Amazon Nova Lite: 26.3 (#259), DeepSeek-V3.1-Terminus: —
| Benchmark | Amazon Nova Lite | DeepSeek-V3.1-Terminus |
|---|---|---|
| Humanity's Last Exam | 3.6% | — |
| MMLU-Pro | 60% | — |
| Vectara Hallucination Rate | 6.1% | — |
| GPQA (HELM) | 39.7% | — |
| LMArena Expert | 1201 | — |
| MMLU | 77% | — |
Multimodal Not comparable
Amazon Nova Lite: 25.5 (#123), DeepSeek-V3.1-Terminus: —
| Benchmark | Amazon Nova Lite | DeepSeek-V3.1-Terminus |
|---|---|---|
| LMArena Vision | 990 | — |
Multilingual DeepSeek-V3.1-Terminus leads
Amazon Nova Lite: 38.0 (#232), DeepSeek-V3.1-Terminus: 52.1 (#92)
| Benchmark | Amazon Nova Lite | DeepSeek-V3.1-Terminus |
|---|---|---|
| LMArena Non-English | 1208 | 1407 |
| LMArena Russian | 1216 | 1436 |
| LMArena Chinese | 1225 | — |
| LMArena French | 1237 | — |
| LMArena German | 1229 | — |
| LMArena Japanese | 1153 | — |
| LMArena Korean | 1153 | — |
| LMArena Spanish | 1226 | — |
Instruction Following DeepSeek-V3.1-Terminus leads
Amazon Nova Lite: 59.3 (#255), DeepSeek-V3.1-Terminus: 74.0 (#106)
| Benchmark | Amazon Nova Lite | DeepSeek-V3.1-Terminus |
|---|---|---|
| LMArena Instruction Following | 1205 | 1404 |
| LiveBench Instruction Following | 54.1% | — |
| IFEval | 77.6% | — |
Long Context DeepSeek-V3.1-Terminus leads
Amazon Nova Lite: 37.4 (#217), DeepSeek-V3.1-Terminus: 43.4 (#97)
| Benchmark | Amazon Nova Lite | DeepSeek-V3.1-Terminus |
|---|---|---|
| LMArena Longer Query | 1234 | 1421 |
Writing & Preference DeepSeek-V3.1-Terminus leads
Amazon Nova Lite: 42.1 (#237), DeepSeek-V3.1-Terminus: 61.0 (#92)
| Benchmark | Amazon Nova Lite | DeepSeek-V3.1-Terminus |
|---|---|---|
| LMArena Text | 1229 | 1419 |
| LMArena Creative Writing | 1197 | 1403 |
| LMArena Multi-Turn | 1199 | 1411 |
| WildBench | 75% | — |
| LiveBench Language | 25.9% | — |
Frequently asked questions
Is Amazon Nova Lite better than DeepSeek-V3.1-Terminus?
DeepSeek-V3.1-Terminus is the stronger model overall, scoring 43.1 to 31.9 on the Noometry Index. Amazon Nova Lite costs 4.3× less per token, which makes it the better buy when DeepSeek-V3.1-Terminus's lead doesn't matter for your workload.
Which is cheaper, Amazon Nova Lite or DeepSeek-V3.1-Terminus?
Amazon Nova Lite is cheaper. It lists at $0.06 per million input tokens and $0.24 per million output tokens; DeepSeek-V3.1-Terminus lists at $0.27 and $1.
Is Amazon Nova Lite or DeepSeek-V3.1-Terminus better for coding?
DeepSeek-V3.1-Terminus scores higher on coding benchmarks: 42.0 versus 32.5 in the Noometry coding category.
Which has the bigger context window?
Amazon Nova Lite does, with 300K tokens against 164K.
How many benchmarks do Amazon Nova Lite and DeepSeek-V3.1-Terminus share?
11 benchmarks have published results for both models. Amazon Nova Lite has 34 scored results on Noometry and DeepSeek-V3.1-Terminus has 16.