Noometry

Model comparison

DeepSeek-V3.1 vs Nova Premier 1.0

DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 38.3 on the Noometry Index.

Last verified . 1 shared benchmarks.

DeepSeek-V3.1 DeepSeek

42.8

Rank #108 Confirmed

Nova Premier 1.0 Amazon

38.3

Rank #189 Confirmed

Summary

  • They share 1 benchmark with published results for both. DeepSeek-V3.1 scores higher in 5 categories and Nova Premier 1.0 in 0 categories; 5 gaps are clear of the uncertainty.
  • The widest gap is in writing & preference, where DeepSeek-V3.1 leads 60.3 to 51.0.
  • The biggest single-benchmark swing is Kagi LLM Benchmark: 53.2% for DeepSeek-V3.1 and 44.8% for Nova Premier 1.0.
  • DeepSeek-V3.1 is cheaper at $0.25 / $0.95 per million input/output tokens, against $2.50 / $12.50 for Nova Premier 1.0.
  • Nova Premier 1.0 accepts more context: 1M tokens versus 164K.
  • DeepSeek-V3.1 has downloadable open weights; the other is API-only.

Side by side

DeepSeek-V3.1Nova Premier 1.0
ProviderDeepSeekAmazon
Noometry Index42.838.3
Released2025-08-212025-04-30
WeightsOpenProprietary
Context window164K1M
Max output8K10K
Input $ / M tokens$0.25$2.50
Output $ / M tokens$0.95$12.50
Results tracked276

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Not comparable

DeepSeek-V3.1: 40.3 (#144), Nova Premier 1.0: —

BenchmarkDeepSeek-V3.1Nova Premier 1.0
WeirdML38.4%—
LMArena Coding1417—

Reasoning DeepSeek-V3.1 leads

DeepSeek-V3.1: 27.9 (#110), Nova Premier 1.0: 23.0 (#185)

BenchmarkDeepSeek-V3.1Nova Premier 1.0
Kagi LLM Benchmark53.2%44.8%
SimpleBench40%—
LMArena Hard Prompts1417—
DTBench82.7%—
LMCA24.3%—
Epoch Capabilities Index139.92—
ForecastBench58—

Math DeepSeek-V3.1 leads

DeepSeek-V3.1: 38.9 (#122), Nova Premier 1.0: 33.8 (#199)

BenchmarkDeepSeek-V3.1Nova Premier 1.0
Omni-MATH—35%
LMArena Math1420—

Knowledge DeepSeek-V3.1 leads

DeepSeek-V3.1: 43.7 (#90), Nova Premier 1.0: 35.8 (#180)

BenchmarkDeepSeek-V3.1Nova Premier 1.0
MMLU-Pro—72.6%
Vectara Hallucination Rate5.5%—
GPQA (HELM)—51.8%
LMArena Expert1405—

Multilingual Not comparable

DeepSeek-V3.1: 51.6 (#106), Nova Premier 1.0: —

BenchmarkDeepSeek-V3.1Nova Premier 1.0
LMArena Non-English1400—
LMArena Chinese1469—
LMArena French1447—
LMArena German1411—
LMArena Japanese1378—
LMArena Korean1337—
LMArena Russian1405—
LMArena Spanish1431—

Instruction Following DeepSeek-V3.1 leads

DeepSeek-V3.1: 73.9 (#110), Nova Premier 1.0: 66.8 (#204)

BenchmarkDeepSeek-V3.1Nova Premier 1.0
IFEval—80.3%
LMArena Instruction Following1400—

Long Context Not comparable

DeepSeek-V3.1: 36.3 (#232), Nova Premier 1.0: —

BenchmarkDeepSeek-V3.1Nova Premier 1.0
Fiction.LiveBench52.8%—
LMArena Longer Query1422—

Writing & Preference DeepSeek-V3.1 leads

DeepSeek-V3.1: 60.3 (#98), Nova Premier 1.0: 51.0 (#178)

BenchmarkDeepSeek-V3.1Nova Premier 1.0
LMArena Text1420—
LMArena Creative Writing1401—
EQ-Bench Creative Writing1436—
WildBench—78.8%
LMArena Multi-Turn1408—

Frequently asked questions

Is DeepSeek-V3.1 better than Nova Premier 1.0?

DeepSeek-V3.1 is the stronger model overall, scoring 42.8 to 38.3 on the Noometry Index.

Which is cheaper, DeepSeek-V3.1 or Nova Premier 1.0?

DeepSeek-V3.1 is cheaper. It lists at $0.25 per million input tokens and $0.95 per million output tokens; Nova Premier 1.0 lists at $2.50 and $12.50.

Which has the bigger context window?

Nova Premier 1.0 does, with 1M tokens against 164K.

How many benchmarks do DeepSeek-V3.1 and Nova Premier 1.0 share?

1 benchmark has published results for both models. DeepSeek-V3.1 has 27 scored results on Noometry and Nova Premier 1.0 has 6.

Related comparisons

Go deeper