Model comparison
DeepSeek V4 Pro vs Kimi K2 Thinking Turbo
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 45.8 on the Noometry Index.
Last verified . 21 shared benchmarks.
DeepSeekSummary
- They share 21 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 8 categories and Kimi K2 Thinking Turbo in 0 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 32.5.
- The biggest single-benchmark swing is Chess Puzzles: 47% for DeepSeek V4 Pro and 20% for Kimi K2 Thinking Turbo.
Side by side
| DeepSeek V4 Pro | Kimi K2 Thinking Turbo | |
|---|---|---|
| Provider | DeepSeek | Moonshot AI |
| Noometry Index | 54.3 | 45.8 |
| Released | 2026-04-24 | 2025-11-06 |
| Weights | Open | Open |
| Context window | 1M | — |
| Max output | 393K | — |
| Input $ / M tokens | $0.66 | — |
| Output $ / M tokens | $1.98 | — |
| Results tracked | 48 | 21 |
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), Kimi K2 Thinking Turbo: 38.4 (#178)
| Benchmark | DeepSeek V4 Pro | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena WebDev | 1582 | 1322 |
| LMArena Coding | 1470 | 1454 |
| SWE-bench Verified | 77.6% | — |
| FrontierCode | 28.6% | — |
| SciCode | 51% | — |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Pro: 32.8 (#58), Kimi K2 Thinking Turbo: —
| Benchmark | DeepSeek V4 Pro | Kimi K2 Thinking Turbo |
|---|---|---|
| APEX-Agents | 47.3% | — |
| Vending-Bench 2 | 3,285 | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Kimi K2 Thinking Turbo: 32.5 (#75)
| Benchmark | DeepSeek V4 Pro | Kimi K2 Thinking Turbo |
|---|---|---|
| Chess Puzzles | 47% | 20% |
| LMArena Hard Prompts | 1461 | 1428 |
| ARC-AGI-2 | 61.3% | — |
| Kagi LLM Benchmark | 53.5% | — |
| NYT Connections (extended) | 91.3% | — |
| ARC-AGI-1 | 90.5% | — |
| CritPt | 18% | — |
| 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), Kimi K2 Thinking Turbo: 47.4 (#68)
| Benchmark | DeepSeek V4 Pro | Kimi K2 Thinking Turbo |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.6% | 83.1% |
| LMArena Math | 1455 | 1429 |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| ProofBench | 50% | — |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), Kimi K2 Thinking Turbo: 50.9 (#69)
| Benchmark | DeepSeek V4 Pro | Kimi K2 Thinking Turbo |
|---|---|---|
| GPQA Diamond | 91.7% | 84.2% |
| LMArena Expert | 1464 | 1439 |
| SimpleQA Verified | 52.9% | — |
| Vectara Hallucination Rate | 8.6% | — |
Multilingual DeepSeek V4 Pro leads
DeepSeek V4 Pro: 54.4 (#45), Kimi K2 Thinking Turbo: 51.4 (#109)
| Benchmark | DeepSeek V4 Pro | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Non-English | 1439 | 1398 |
| LMArena Chinese | 1486 | 1456 |
| LMArena French | 1472 | 1425 |
| LMArena German | 1458 | 1390 |
| LMArena Japanese | 1445 | 1357 |
| LMArena Korean | 1447 | 1330 |
| LMArena Russian | 1453 | 1391 |
| LMArena Spanish | 1458 | 1406 |
Instruction Following DeepSeek V4 Pro leads
DeepSeek V4 Pro: 76.1 (#47), Kimi K2 Thinking Turbo: 74.0 (#109)
| Benchmark | DeepSeek V4 Pro | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Instruction Following | 1448 | 1403 |
Long Context DeepSeek V4 Pro leads
DeepSeek V4 Pro: 45.0 (#51), Kimi K2 Thinking Turbo: 43.2 (#102)
| Benchmark | DeepSeek V4 Pro | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Longer Query | 1458 | 1415 |
| CL-bench Life | 13.5% | — |
Writing & Preference DeepSeek V4 Pro leads
DeepSeek V4 Pro: 65.5 (#46), Kimi K2 Thinking Turbo: 60.0 (#104)
| Benchmark | DeepSeek V4 Pro | Kimi K2 Thinking Turbo |
|---|---|---|
| LMArena Text | 1451 | 1415 |
| LMArena Creative Writing | 1446 | 1374 |
| LMArena Multi-Turn | 1467 | 1414 |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Kimi K2 Thinking Turbo?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 45.8 on the Noometry Index.
Is DeepSeek V4 Pro or Kimi K2 Thinking Turbo better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 38.4 in the Noometry coding category.
How many benchmarks do DeepSeek V4 Pro and Kimi K2 Thinking Turbo share?
21 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Kimi K2 Thinking Turbo has 21.