Model comparison
QwQ-32B vs Tulu 3 (Tülu 3) 70B
QwQ-32B is the stronger model overall, scoring 39.4 to 32.6 on the Noometry Index.
Last verified . 13 shared benchmarks.
Alibaba (Qwen)
Allen Institute for AI (Ai2)Summary
- They share 13 benchmarks with published results for both. QwQ-32B scores higher in 7 categories and Tulu 3 (Tülu 3) 70B in 1 category; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where QwQ-32B leads 36.2 to 13.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 59.2% for QwQ-32B and 4.4% for Tulu 3 (Tülu 3) 70B.
Side by side
| QwQ-32B | Tulu 3 (Tülu 3) 70B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Allen Institute for AI (Ai2) |
| Noometry Index | 39.4 | 32.6 |
| Released | 2024-11-28 | 2024-11-21 |
| Weights | Open | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 36 | 14 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
QwQ-32B: 34.9 (#220), Tulu 3 (Tülu 3) 70B: 35.4 (#211)
| Benchmark | QwQ-32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Coding | 1333 | 1236 |
| Aider Polyglot | 20.9% | — |
| BigCodeBench Instruct | 44.6% | — |
| LiveBench Coding | 72.2% | — |
| BigCodeBench Complete | 54.4% | — |
Reasoning Too close to call
QwQ-32B: 22.9 (#171), Tulu 3 (Tülu 3) 70B: 22.9 (#172)
| Benchmark | QwQ-32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Hard Prompts | 1326 | 1220 |
| Chess Puzzles | 5% | — |
| LiveBench Reasoning | 83.5% | — |
| LiveBench Data Analysis | 65% | — |
| Epoch Capabilities Index | 137.6 | — |
| ForecastBench | 58.3 | — |
| LiveBench | 72% | — |
Math QwQ-32B leads
QwQ-32B: 36.2 (#123), Tulu 3 (Tülu 3) 70B: 13.3 (#297)
| Benchmark | QwQ-32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 59.2% | 4.4% |
| LMArena Math | 1359 | 1242 |
| LiveBench Math | 77.8% | — |
| MATH Level 5 | — | 42.7% |
Knowledge QwQ-32B leads
QwQ-32B: 37.2 (#150), Tulu 3 (Tülu 3) 70B: 25.1 (#257)
| Benchmark | QwQ-32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| GPQA Diamond | 65.3% | 46.3% |
| Confabulations | 15.6% | — |
| LMArena Expert | 1323 | — |
Multilingual QwQ-32B leads
QwQ-32B: 45.0 (#171), Tulu 3 (Tülu 3) 70B: 40.1 (#217)
| Benchmark | QwQ-32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Non-English | 1306 | 1236 |
| LMArena Chinese | 1378 | 1249 |
| LMArena Russian | 1297 | 1246 |
| LMArena French | 1337 | — |
| LMArena German | 1313 | — |
| LMArena Japanese | 1263 | — |
| LMArena Korean | 1278 | — |
| LMArena Spanish | 1353 | — |
Instruction Following QwQ-32B leads
QwQ-32B: 72.8 (#131), Tulu 3 (Tülu 3) 70B: 65.0 (#222)
| Benchmark | QwQ-32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Instruction Following | 1297 | 1233 |
| LiveBench Instruction Following | 81.8% | — |
Long Context QwQ-32B leads
QwQ-32B: 49.0 (#11), Tulu 3 (Tülu 3) 70B: 37.1 (#217)
| Benchmark | QwQ-32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Longer Query | 1308 | 1224 |
| Fiction.LiveBench | 83.3% | — |
Writing & Preference QwQ-32B leads
QwQ-32B: 50.9 (#175), Tulu 3 (Tülu 3) 70B: 45.8 (#218)
| Benchmark | QwQ-32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Text | 1329 | 1256 |
| LMArena Creative Writing | 1288 | 1231 |
| LMArena Multi-Turn | 1314 | 1252 |
| Short-Story Creative Writing | 80.2% | — |
| EQ-Bench Creative Writing | 1257 | — |
| LiveBench Language | 51.4% | — |
Frequently asked questions
Is QwQ-32B better than Tulu 3 (Tülu 3) 70B?
QwQ-32B is the stronger model overall, scoring 39.4 to 32.6 on the Noometry Index.
Is QwQ-32B or Tulu 3 (Tülu 3) 70B better for coding?
They score almost the same on coding (34.9 vs 35.4); test both on your own repository before choosing.
How many benchmarks do QwQ-32B and Tulu 3 (Tülu 3) 70B share?
13 benchmarks have published results for both models. QwQ-32B has 36 scored results on Noometry and Tulu 3 (Tülu 3) 70B has 14.