Model comparison
DeepSeek V4 Pro vs Qwen3.6 Max Preview
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 51.5 on the Noometry Index.
Last verified . 26 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 26 benchmarks with published results for both. DeepSeek V4 Pro scores higher in 8 categories and Qwen3.6 Max Preview in 0 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where DeepSeek V4 Pro leads 56.5 to 41.7.
- The biggest single-benchmark swing is Chess Puzzles: 47% for DeepSeek V4 Pro and 20% for Qwen3.6 Max Preview.
- DeepSeek V4 Pro is cheaper at $0.66 / $1.98 per million input/output tokens, against $1.30 / $7.80 for Qwen3.6 Max Preview.
- DeepSeek V4 Pro accepts more context: 1M tokens versus 262K.
- DeepSeek V4 Pro has downloadable open weights; the other is API-only.
Side by side
| DeepSeek V4 Pro | Qwen3.6 Max Preview | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 54.3 | 51.5 |
| Released | 2026-04-24 | 2026-04-20 |
| Weights | Open | Proprietary |
| Context window | 1M | 262K |
| Max output | 393K | 66K |
| Input $ / M tokens | $0.66 | $1.30 |
| Output $ / M tokens | $1.98 | $7.80 |
| Results tracked | 48 | 29 |
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), Qwen3.6 Max Preview: 48.7 (#54)
| Benchmark | DeepSeek V4 Pro | Qwen3.6 Max Preview |
|---|---|---|
| SWE-bench Verified | 77.6% | 76.7% |
| LMArena WebDev | 1582 | 1482 |
| LMArena Coding | 1470 | 1471 |
| FrontierCode | 28.6% | — |
| SciCode | 51% | — |
| WeirdML | 66.2% | — |
| ALE-Bench | 1,403 | — |
Agentic & Tool Use Not comparable
DeepSeek V4 Pro: 32.8 (#58), Qwen3.6 Max Preview: —
| Benchmark | DeepSeek V4 Pro | Qwen3.6 Max Preview |
|---|---|---|
| Vending-Bench 2 | 3,285 | 4,254 |
| APEX-Agents | 47.3% | — |
Reasoning DeepSeek V4 Pro leads
DeepSeek V4 Pro: 56.5 (#24), Qwen3.6 Max Preview: 41.7 (#53)
| Benchmark | DeepSeek V4 Pro | Qwen3.6 Max Preview |
|---|---|---|
| NYT Connections (extended) | 91.3% | 74.1% |
| Chess Puzzles | 47% | 20% |
| LMArena Hard Prompts | 1461 | 1457 |
| Mystery Game Puzzles | 43% | 19% |
| DTBench | 93.9% | 87.2% |
| LMCA | 45.5% | 42.5% |
| Epoch Capabilities Index | 155.31 | 149.24 |
| ARC-AGI-2 | 61.3% | — |
| SimpleBench | — | 63% |
| Kagi LLM Benchmark | 53.5% | — |
| ARC-AGI-1 | 90.5% | — |
| CritPt | 18% | — |
| Surface Evolver Bench | 40% | — |
| ForecastBench | 56.1 | — |
Math DeepSeek V4 Pro leads
DeepSeek V4 Pro: 64.8 (#30), Qwen3.6 Max Preview: 54.1 (#46)
| Benchmark | DeepSeek V4 Pro | Qwen3.6 Max Preview |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 98.6% | 91.1% |
| LMArena Math | 1455 | 1465 |
| FrontierMath (Tiers 1-3) | 64.6% | — |
| FrontierMath Tier 4 | 26.8% | — |
| MathArena Final-Answer Competitions | 76.6% | — |
| ProofBench | 50% | — |
| FrontierMath (Feb 2025 set) | — | 23.1% |
| FrontierMath Tier 4 (v1) | — | 4.2% |
Knowledge DeepSeek V4 Pro leads
DeepSeek V4 Pro: 59.5 (#31), Qwen3.6 Max Preview: 57.6 (#39)
| Benchmark | DeepSeek V4 Pro | Qwen3.6 Max Preview |
|---|---|---|
| GPQA Diamond | 91.7% | 87.4% |
| SimpleQA Verified | 52.9% | 52% |
| LMArena Expert | 1464 | 1478 |
| Vectara Hallucination Rate | 8.6% | — |
Multilingual Too close to call
DeepSeek V4 Pro: 54.4 (#45), Qwen3.6 Max Preview: 54.2 (#48)
| Benchmark | DeepSeek V4 Pro | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Non-English | 1439 | 1437 |
| LMArena Chinese | 1486 | 1487 |
| LMArena French | 1472 | 1449 |
| LMArena Russian | 1453 | 1445 |
| LMArena Spanish | 1458 | 1454 |
| LMArena German | 1458 | — |
| LMArena Japanese | 1445 | — |
| LMArena Korean | 1447 | — |
Instruction Following Too close to call
DeepSeek V4 Pro: 76.1 (#47), Qwen3.6 Max Preview: 75.7 (#55)
| Benchmark | DeepSeek V4 Pro | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Instruction Following | 1448 | 1438 |
Long Context Too close to call
DeepSeek V4 Pro: 45.0 (#51), Qwen3.6 Max Preview: 44.6 (#61)
| Benchmark | DeepSeek V4 Pro | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Longer Query | 1458 | 1457 |
| CL-bench Life | 13.5% | — |
Writing & Preference DeepSeek V4 Pro leads
DeepSeek V4 Pro: 65.5 (#46), Qwen3.6 Max Preview: 63.8 (#60)
| Benchmark | DeepSeek V4 Pro | Qwen3.6 Max Preview |
|---|---|---|
| LMArena Text | 1451 | 1447 |
| LMArena Creative Writing | 1446 | 1435 |
| LMArena Multi-Turn | 1467 | 1456 |
| EQ-Bench Creative Writing | 1553 | — |
| EQ-Bench 4 | 1166 | — |
Frequently asked questions
Is DeepSeek V4 Pro better than Qwen3.6 Max Preview?
DeepSeek V4 Pro is the stronger model overall, scoring 54.3 to 51.5 on the Noometry Index.
Which is cheaper, DeepSeek V4 Pro or Qwen3.6 Max Preview?
DeepSeek V4 Pro is cheaper. It lists at $0.66 per million input tokens and $1.98 per million output tokens; Qwen3.6 Max Preview lists at $1.30 and $7.80.
Is DeepSeek V4 Pro or Qwen3.6 Max Preview better for coding?
DeepSeek V4 Pro scores higher on coding benchmarks: 52.4 versus 48.7 in the Noometry coding category.
Which has the bigger context window?
DeepSeek V4 Pro does, with 1M tokens against 262K.
How many benchmarks do DeepSeek V4 Pro and Qwen3.6 Max Preview share?
26 benchmarks have published results for both models. DeepSeek V4 Pro has 48 scored results on Noometry and Qwen3.6 Max Preview has 29.