Model comparison
DeepSeek V4 Pro vs Nova 2 Lite
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 39.7 on the Noometry Index.
Last verified . 18 shared benchmarks.
DeepSeekSummary
- They share 18 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 9 categories and Nova 2 Lite in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 27.5.
- Nova 2 Lite is cheaper at $0.30 / $2.50 per million input/output tokens, against $0.66 / $1.98 for DeepSeek V4 Pro.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4 Pro | Nova 2 Lite | |
|---|---|---|
| Provider | DeepSeek | Amazon |
| Noometry Index | 54.3 | 39.7 |
| Released | 2026-04-24 | 2025-12-01 |
| Weights | Open | Proprietary |
| Context window | 1M | 1M |
| Max output | 393K | 64K |
| Input $ / M tokens | $0.66 | $0.30 |
| Output $ / M tokens | $1.98 | $2.50 |
| Results tracked | 48 | 19 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding DeepSeek V4 Pro leads
DeepSeek V4 Pro: 52.4 (#34), Nova 2 Lite: 40.7 (#134)
| Benchmark | DeepSeek V4 Pro | Nova 2 Lite |
|---|---|---|
| LMArena Coding | 1470 | 1385 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| LMArena WebDev | 1582 | — |
| SciCode | 51% | — |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use DeepSeek V4 Pro leads
DeepSeek V4 Pro: 32.8 (#58), Nova 2 Lite: 24.1 (#121)
| Benchmark | DeepSeek V4 Pro | Nova 2 Lite |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Berkeley Function Calling Leaderboard | — | 27.1% |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Nova 2 Lite: 27.5 (#118)
| Benchmark | DeepSeek V4 Pro | Nova 2 Lite |
|---|---|---|
| LMArena Hard Prompts | 1461 | 1364 |
| 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% | — |
| Mystery Game Puzzles | 43% | — |
| DTBench | 93.9% | — |
| LMCA | 45.5% | — |
| Surface Evolver Bench | 40% | — |
| Epoch Capabilities Index | 155.31 | — |
| ForecastBench | 56.1 | — |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), Nova 2 Lite: 37.5 (#156)
| Benchmark | DeepSeek V4 Pro | Nova 2 Lite |
|---|---|---|
| LMArena Math | 1455 | 1359 |
| 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% | — |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), Nova 2 Lite: 43.0 (#94)
| Benchmark | DeepSeek V4 Pro | Nova 2 Lite |
|---|---|---|
| Vectara Hallucination Rate | 8.6% | 5.1% |
| LMArena Expert | 1464 | 1358 |
| GPQA Diamond | 91.7% | — |
| SimpleQA Verified | 52.9% | — |
Multilingual DeepSeek V4 Pro leads
DeepSeek V4 Pro: 54.4 (#45), Nova 2 Lite: 47.1 (#153)
| Benchmark | DeepSeek V4 Pro | Nova 2 Lite |
|---|---|---|
| LMArena Non-English | 1439 | 1337 |
| LMArena Chinese | 1486 | 1364 |
| LMArena French | 1472 | 1381 |
| LMArena German | 1458 | 1343 |
| LMArena Japanese | 1445 | 1271 |
| LMArena Korean | 1447 | 1284 |
| LMArena Russian | 1453 | 1343 |
| LMArena Spanish | 1458 | 1373 |
Instruction Following DeepSeek V4 Pro leads
DeepSeek V4 Pro: 76.1 (#47), Nova 2 Lite: 70.5 (#161)
| Benchmark | DeepSeek V4 Pro | Nova 2 Lite |
|---|---|---|
| LMArena Instruction Following | 1448 | 1335 |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#51), Nova 2 Lite: 40.6 (#150)
| Benchmark | DeepSeek V4 Pro | Nova 2 Lite |
|---|---|---|
| LMArena Longer Query | 1458 | 1335 |
| CL-bench Life | 13.5% | — |
Writing & Preference DeepSeek V4 Pro leads
DeepSeek V4 Pro: 65.5 (#46), Nova 2 Lite: 53.9 (#154)
| Benchmark | DeepSeek V4 Pro | Nova 2 Lite |
|---|---|---|
| LMArena Text | 1451 | 1362 |
| LMArena Creative Writing | 1446 | 1291 |
| LMArena Multi-Turn | 1467 | 1338 |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Nova 2 Lite?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 39.7 on the Noometry Index.
Which is cheaper, DeepSeek V4 Pro or Nova 2 Lite?
Nova 2 Lite is cheaper. It lists at $0.30 per million input tokens and $2.50 per million output tokens; DeepSeek V4 Pro lists at $0.66 and $1.98.
Is DeepSeek V4 Pro or Nova 2 Lite better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 40.7 in the Noometry coding category.
Which has the bigger context window?
Both accept 1M tokens.
How many benchmarks do DeepSeek V4 Pro and Nova 2 Lite share?
18 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Nova 2 Lite has 19.