Model comparison
Qwen3 32B vs Tulu 3 (Tülu 3) 70B
Qwen3 32B is the stronger model overall, scoring 39.2 to 33.0 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. Qwen3 32B scores higher in 7 categories and Tulu 3 (Tülu 3) 70B in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3 32B leads 39.7 to 14.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 66.9% for Qwen3 32B and 4.4% for Tulu 3 (Tülu 3) 70B.
Side by side
| Qwen3 32B | Tulu 3 (Tülu 3) 70B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Allen Institute for AI (Ai2) |
| Noometry Index | 39.2 | 33.0 |
| Released | 2025-04 | 2024-11-21 |
| Weights | Open | Open |
| Context window | 131K | — |
| Max output | 16K | — |
| Input $ / M tokens | $0.70 | — |
| Output $ / M tokens | $2.80 | — |
| Results tracked | 26 | 14 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 32B leads
Qwen3 32B: 37.7 (#190), Tulu 3 (Tülu 3) 70B: 36.0 (#214)
| Benchmark | Qwen3 32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Coding | 1358 | 1235 |
| Aider Polyglot | 40% | — |
| SciCode | 35.4% | — |
Agentic & Tool Use Not comparable
Qwen3 32B: 32.6 (#62), Tulu 3 (Tülu 3) 70B: —
| Benchmark | Qwen3 32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 48.7% | — |
Reasoning Tulu 3 (Tülu 3) 70B leads
Qwen3 32B: 20.2 (#241), Tulu 3 (Tülu 3) 70B: 23.9 (#169)
| Benchmark | Qwen3 32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Hard Prompts | 1334 | 1220 |
| Kagi LLM Benchmark | 54.9% | — |
| CritPt | 0.3% | — |
| Chess Puzzles | 5% | — |
| DTBench | 67.5% | — |
| LMCA | 17.3% | — |
| Epoch Capabilities Index | 138.51 | — |
Math Qwen3 32B leads
Qwen3 32B: 39.7 (#99), Tulu 3 (Tülu 3) 70B: 14.2 (#303)
| Benchmark | Qwen3 32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 66.9% | 4.4% |
| LMArena Math | 1399 | 1242 |
| MATH Level 5 | — | 42.7% |
Knowledge Qwen3 32B leads
Qwen3 32B: 40.0 (#125), Tulu 3 (Tülu 3) 70B: 25.0 (#264)
| Benchmark | Qwen3 32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| GPQA Diamond | 65.7% | 46.3% |
| Vectara Hallucination Rate | 5.9% | — |
| LMArena Expert | 1362 | — |
Multilingual Qwen3 32B leads
Qwen3 32B: 45.6 (#167), Tulu 3 (Tülu 3) 70B: 39.9 (#222)
| Benchmark | Qwen3 32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Non-English | 1317 | 1236 |
| LMArena Chinese | 1357 | 1249 |
| LMArena Russian | 1311 | 1246 |
| LMArena German | 1341 | — |
Instruction Following Qwen3 32B leads
Qwen3 32B: 68.9 (#179), Tulu 3 (Tülu 3) 70B: 64.8 (#227)
| Benchmark | Qwen3 32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Instruction Following | 1305 | 1233 |
Long Context Qwen3 32B leads
Qwen3 32B: 43.8 (#87), Tulu 3 (Tülu 3) 70B: 37.1 (#222)
| Benchmark | Qwen3 32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Longer Query | 1327 | 1224 |
| Fiction.LiveBench | 74.2% | — |
Writing & Preference Qwen3 32B leads
Qwen3 32B: 52.9 (#163), Tulu 3 (Tülu 3) 70B: 45.6 (#223)
| Benchmark | Qwen3 32B | Tulu 3 (Tülu 3) 70B |
|---|---|---|
| LMArena Text | 1340 | 1256 |
| LMArena Creative Writing | 1297 | 1231 |
| LMArena Multi-Turn | 1331 | 1252 |
Frequently asked questions
Is Qwen3 32B better than Tulu 3 (Tülu 3) 70B?
Qwen3 32B is the stronger model overall, scoring 39.2 to 33.0 on the Noometry Index.
Is Qwen3 32B or Tulu 3 (Tülu 3) 70B better for coding?
Qwen3 32B scores higher on coding benchmarks: 37.7 versus 36.0 in the Noometry coding category.
How many benchmarks do Qwen3 32B and Tulu 3 (Tülu 3) 70B share?
13 benchmarks have published results for both models. Qwen3 32B has 26 scored results on Noometry and Tulu 3 (Tülu 3) 70B has 14.