Model comparison
DeepSeek-R1 vs Qwen3.8 Max
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 42.3 on the Noometry Index. DeepSeek-R1 costs 3.3× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Last verified . 22 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 22 benchmarks with published results for both. DeepSeek-R1 scores higher in 0 categories and Qwen3.8 Max in 9 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Qwen3.8 Max leads 54.4 to 18.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 66.4% for DeepSeek-R1 and 100% for Qwen3.8 Max.
- DeepSeek-R1 is cheaper at $0.50 / $2.15 per million input/output tokens, against $2 / $6 for Qwen3.8 Max.
- Qwen3.8 Max accepts more context: 1M tokens versus 164K.
Side by side
| DeepSeek-R1 | Qwen3.8 Max | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 42.3 | 56.8 |
| Released | 2025-01-20 | 2026-08-02 |
| Weights | Proprietary | Proprietary |
| Context window | 164K | 1M |
| Max output | 64K | 131K |
| Input $ / M tokens | $0.50 | $2 |
| Output $ / M tokens | $2.15 | $6 |
| Results tracked | 52 | 39 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3.8 Max leads
DeepSeek-R1: 46.3 (#68), Qwen3.8 Max: 53.5 (#29)
| Benchmark | DeepSeek-R1 | Qwen3.8 Max |
|---|---|---|
| SciCode | 35.7% | 53.2% |
| LMArena Coding | 1427 | 1502 |
| DeepSWE | — | 57.5% |
| Aider Polyglot | 71.4% | — |
| LMArena WebDev | — | 1674 |
| FrontierSWE | — | 17.8% |
| WeirdML | 41.6% | — |
| LiveBench Coding | 66.7% | — |
| ALE-Bench | 804.12 | — |
| AlgoTune | 1.7 | — |
Agentic & Tool Use Qwen3.8 Max leads
DeepSeek-R1: 30.7 (#75), Qwen3.8 Max: 45.4 (#14)
| Benchmark | DeepSeek-R1 | Qwen3.8 Max |
|---|---|---|
| APEX-Agents | — | 63.3% |
| τ²-bench Banking | — | 55.1% |
| DeepResearch Bench | 35.1% | — |
| BALROG | 34.9% | — |
| GDP.pdf | — | 23.2% |
| METR Time Horizons | 53.8% | — |
Reasoning Qwen3.8 Max leads
DeepSeek-R1: 18.6 (#278), Qwen3.8 Max: 54.4 (#26)
| Benchmark | DeepSeek-R1 | Qwen3.8 Max |
|---|---|---|
| CritPt | 1.1% | 20% |
| LMArena Hard Prompts | 1416 | 1496 |
| Epoch Capabilities Index | 141.29 | 156.41 |
| ARC-AGI-2 | 1.3% | — |
| SimpleBench | 40.8% | — |
| Kagi LLM Benchmark | 69.4% | — |
| NYT Connections (extended) | — | 88.3% |
| ARC-AGI-1 | 21.2% | — |
| Chess Puzzles | — | 40% |
| LiveBench Reasoning | 83.2% | — |
| Mystery Game Puzzles | — | 38% |
| DTBench | — | 92% |
| LiveBench Data Analysis | 69.8% | — |
| LMCA | — | 46.2% |
| ForecastBench | 60 | — |
| LiveBench | 71.6% | — |
Math Qwen3.8 Max leads
DeepSeek-R1: 43.8 (#79), Qwen3.8 Max: 73.2 (#20)
| Benchmark | DeepSeek-R1 | Qwen3.8 Max |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.4% | 100% |
| LMArena Math | 1400 | 1499 |
| FrontierMath (Tiers 1-3) | — | 74.7% |
| FrontierMath Tier 4 | — | 46.3% |
| ProofBench | — | 58% |
| Omni-MATH | 42.4% | — |
| LiveBench Math | 80.7% | — |
| MATH Level 5 | 96.6% | — |
Knowledge Qwen3.8 Max leads
DeepSeek-R1: 44.5 (#87), Qwen3.8 Max: 61.7 (#27)
| Benchmark | DeepSeek-R1 | Qwen3.8 Max |
|---|---|---|
| GPQA Diamond | 76.3% | 92.7% |
| LMArena Expert | 1394 | 1507 |
| SimpleQA Verified | — | 47.3% |
| MMLU-Pro | 79.3% | — |
| Confabulations | 12.7% | — |
| Vectara Hallucination Rate | 11.3% | — |
| GPQA (HELM) | 66.6% | — |
Multimodal Not comparable
DeepSeek-R1: —, Qwen3.8 Max: 37.2 (#75)
| Benchmark | DeepSeek-R1 | Qwen3.8 Max |
|---|---|---|
| LMArena Vision | — | 1314 |
| Furniture Assembly | — | 20% |
Multilingual Qwen3.8 Max leads
DeepSeek-R1: 52.4 (#85), Qwen3.8 Max: 56.7 (#18)
| Benchmark | DeepSeek-R1 | Qwen3.8 Max |
|---|---|---|
| LMArena Non-English | 1412 | 1472 |
| LMArena Chinese | 1442 | 1538 |
| LMArena French | 1417 | 1503 |
| LMArena German | 1404 | 1483 |
| LMArena Japanese | 1391 | 1467 |
| LMArena Korean | 1360 | 1461 |
| LMArena Russian | 1423 | 1481 |
| LMArena Spanish | 1411 | 1492 |
Instruction Following Qwen3.8 Max leads
DeepSeek-R1: 72.0 (#143), Qwen3.8 Max: 77.6 (#17)
| Benchmark | DeepSeek-R1 | Qwen3.8 Max |
|---|---|---|
| LMArena Instruction Following | 1382 | 1479 |
| LiveBench Instruction Following | 80.5% | — |
| IFEval | 78.4% | — |
Long Context Too close to call
DeepSeek-R1: 45.4 (#36), Qwen3.8 Max: 45.6 (#31)
| Benchmark | DeepSeek-R1 | Qwen3.8 Max |
|---|---|---|
| LMArena Longer Query | 1391 | 1489 |
| Fiction.LiveBench | 75% | — |
Writing & Preference Qwen3.8 Max leads
DeepSeek-R1: 61.4 (#88), Qwen3.8 Max: 67.1 (#30)
| Benchmark | DeepSeek-R1 | Qwen3.8 Max |
|---|---|---|
| LMArena Text | 1428 | 1483 |
| LMArena Creative Writing | 1405 | 1479 |
| LMArena Multi-Turn | 1405 | 1489 |
| Short-Story Creative Writing | 83% | — |
| EQ-Bench Creative Writing | 1500 | — |
| WildBench | 82.8% | — |
| LiveBench Language | 48.5% | — |
Frequently asked questions
Is DeepSeek-R1 better than Qwen3.8 Max?
Qwen3.8 Max is the stronger model overall, scoring 56.8 to 42.3 on the Noometry Index. DeepSeek-R1 costs 3.3× less per token, which makes it the better buy when Qwen3.8 Max's lead doesn't matter for your workload.
Which is cheaper, DeepSeek-R1 or Qwen3.8 Max?
DeepSeek-R1 is cheaper. It lists at $0.50 per million input tokens and $2.15 per million output tokens; Qwen3.8 Max lists at $2 and $6.
Is DeepSeek-R1 or Qwen3.8 Max better for coding?
Qwen3.8 Max scores higher on coding benchmarks: 53.5 versus 46.3 in the Noometry coding category.
Which has the bigger context window?
Qwen3.8 Max does, with 1M tokens against 164K.
How many benchmarks do DeepSeek-R1 and Qwen3.8 Max share?
22 benchmarks have published results for both models. DeepSeek-R1 has 52 scored results on Noometry and Qwen3.8 Max has 39.