Model comparison
DeepSeek LLM 67B vs Qwen3-30B-A3B
Qwen3-30B-A3B is the stronger model overall, scoring 38.9 to 24.9 on the Noometry Index.
Last verified . 14 shared benchmarks.
DeepSeek
Alibaba (Qwen)Summary
- They share 14 benchmarks with published results for both. DeepSeek LLM 67B scores higher in 1 category and Qwen3-30B-A3B in 7 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3-30B-A3B leads 41.8 to 7.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 0.8% for DeepSeek LLM 67B and 70.3% for Qwen3-30B-A3B.
Side by side
| DeepSeek LLM 67B | Qwen3-30B-A3B | |
|---|---|---|
| Provider | DeepSeek | Alibaba (Qwen) |
| Noometry Index | 24.9 | 38.9 |
| Released | 2023-11-29 | 2025-04-28 |
| Weights | Open | Open |
| Context window | — | 41K |
| Max output | — | 16K |
| Input $ / M tokens | — | $0.12 |
| Output $ / M tokens | — | $0.50 |
| Results tracked | 15 | 32 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3-30B-A3B leads
DeepSeek LLM 67B: 31.9 (#278), Qwen3-30B-A3B: 37.5 (#194)
| Benchmark | DeepSeek LLM 67B | Qwen3-30B-A3B |
|---|---|---|
| LMArena Coding | 1096 | 1416 |
| SciCode | — | 33.3% |
| WeirdML | — | 29.8% |
Agentic & Tool Use Not comparable
DeepSeek LLM 67B: —, Qwen3-30B-A3B: 29.8 (#82)
| Benchmark | DeepSeek LLM 67B | Qwen3-30B-A3B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 41.4% |
Reasoning Qwen3-30B-A3B leads
DeepSeek LLM 67B: 16.5 (#304), Qwen3-30B-A3B: 22.2 (#204)
| Benchmark | DeepSeek LLM 67B | Qwen3-30B-A3B |
|---|---|---|
| Chess Puzzles | 0% | 8% |
| LMArena Hard Prompts | 1070 | 1398 |
| Epoch Capabilities Index | 110.5 | 139.63 |
| Kagi LLM Benchmark | — | 54.9% |
| CritPt | — | 0.3% |
| DTBench | — | 69.3% |
| LMCA | — | 22.4% |
Math Qwen3-30B-A3B leads
DeepSeek LLM 67B: 8.7 (#324), Qwen3-30B-A3B: 37.4 (#157)
| Benchmark | DeepSeek LLM 67B | Qwen3-30B-A3B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 0.8% | 70.3% |
| LMArena Math | 1108 | 1394 |
| MathArena Final-Answer Competitions | — | 47.8% |
| MATH Level 5 | 6.4% | — |
Knowledge Qwen3-30B-A3B leads
DeepSeek LLM 67B: 7.0 (#313), Qwen3-30B-A3B: 41.8 (#105)
| Benchmark | DeepSeek LLM 67B | Qwen3-30B-A3B |
|---|---|---|
| GPQA Diamond | 24.6% | 70.1% |
| Confabulations | — | 12.3% |
| LMArena Expert | — | 1396 |
Multilingual Qwen3-30B-A3B leads
DeepSeek LLM 67B: 29.4 (#267), Qwen3-30B-A3B: 49.5 (#132)
| Benchmark | DeepSeek LLM 67B | Qwen3-30B-A3B |
|---|---|---|
| LMArena Non-English | 1073 | 1372 |
| LMArena Chinese | 1132 | 1433 |
| LMArena French | — | 1418 |
| LMArena German | — | 1380 |
| LMArena Japanese | — | 1337 |
| LMArena Korean | — | 1331 |
| LMArena Russian | — | 1370 |
| LMArena Spanish | — | 1404 |
Instruction Following Qwen3-30B-A3B leads
DeepSeek LLM 67B: 55.4 (#277), Qwen3-30B-A3B: 72.0 (#142)
| Benchmark | DeepSeek LLM 67B | Qwen3-30B-A3B |
|---|---|---|
| LMArena Instruction Following | 1079 | 1363 |
Long Context DeepSeek LLM 67B leads
DeepSeek LLM 67B: 33.1 (#265), Qwen3-30B-A3B: 31.0 (#283)
| Benchmark | DeepSeek LLM 67B | Qwen3-30B-A3B |
|---|---|---|
| LMArena Longer Query | 1092 | 1379 |
| Fiction.LiveBench | — | 40.6% |
Writing & Preference Qwen3-30B-A3B leads
DeepSeek LLM 67B: 31.6 (#282), Qwen3-30B-A3B: 55.6 (#143)
| Benchmark | DeepSeek LLM 67B | Qwen3-30B-A3B |
|---|---|---|
| LMArena Text | 1105 | 1384 |
| LMArena Creative Writing | 1067 | 1317 |
| LMArena Multi-Turn | 1082 | 1378 |
| Short-Story Creative Writing | — | 75.3% |
Frequently asked questions
Is DeepSeek LLM 67B better than Qwen3-30B-A3B?
Qwen3-30B-A3B is the stronger model overall, scoring 38.9 to 24.9 on the Noometry Index.
Is DeepSeek LLM 67B or Qwen3-30B-A3B better for coding?
Qwen3-30B-A3B scores higher on coding benchmarks: 37.5 versus 31.9 in the Noometry coding category.
How many benchmarks do DeepSeek LLM 67B and Qwen3-30B-A3B share?
14 benchmarks have published results for both models. DeepSeek LLM 67B has 15 scored results on Noometry and Qwen3-30B-A3B has 32.