Model comparison
Qwen2.5 72B Instruct vs Qwen3-4B
Qwen2.5 72B Instruct and Qwen3-4B score almost the same on the Noometry Index (31.9 vs 31.9), so choose on price, context window or the category you care about most.
Last verified . 2 shared benchmarks.
Alibaba (Qwen)Summary
- They share 2 benchmarks with published results for both. Qwen2.5 72B Instruct scores higher in 1 category and Qwen3-4B in 3 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3-4B leads 29.7 to 19.3.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 8.1% for Qwen2.5 72B Instruct and 52.2% for Qwen3-4B.
Side by side
| Qwen2.5 72B Instruct | Qwen3-4B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 31.9 | 31.9 |
| Released | 2024-09 | 2025-04-29 |
| Weights | Open | Open |
| Context window | 131K | — |
| Max output | 8K | — |
| Input $ / M tokens | $1.40 | — |
| Output $ / M tokens | $5.60 | — |
| Results tracked | 43 | 6 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Qwen2.5 72B Instruct: 33.2 (#260), Qwen3-4B: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3-4B |
|---|---|---|
| WeirdML | 16% | — |
| BigCodeBench Instruct | 45.8% | — |
| LMArena Coding | 1292 | — |
| BigCodeBench Complete | 55.9% | — |
Agentic & Tool Use Qwen3-4B leads
Qwen2.5 72B Instruct: 22.1 (#133), Qwen3-4B: 27.6 (#100)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-4B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 35.7% |
| TheAgentCompany | 5.7% | — |
| BALROG | 16.2% | — |
| METR Time Horizons | 35.8% | — |
Reasoning Qwen2.5 72B Instruct leads
Qwen2.5 72B Instruct: 22.3 (#199), Qwen3-4B: 19.2 (#268)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-4B |
|---|---|---|
| Chess Puzzles | — | 4% |
| LMArena Hard Prompts | 1271 | — |
| DTBench | 62.9% | — |
| LMCA | 13.4% | — |
| BIG-Bench Hard | 79.8% | — |
| Epoch Capabilities Index | 129 | — |
| ForecastBench | 57.5 | — |
| HellaSwag | 84.8% | — |
| PIQA | 82.6% | — |
| WinoGrande | 82.3% | — |
Math Qwen3-4B leads
Qwen2.5 72B Instruct: 19.3 (#287), Qwen3-4B: 29.7 (#240)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-4B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 8.1% | 52.2% |
| MathArena Final-Answer Competitions | — | 38.5% |
| Omni-MATH | 33% | — |
| LMArena Math | 1283 | — |
| MATH Level 5 | 63.2% | — |
Knowledge Qwen3-4B leads
Qwen2.5 72B Instruct: 27.0 (#253), Qwen3-4B: 33.0 (#208)
| Benchmark | Qwen2.5 72B Instruct | Qwen3-4B |
|---|---|---|
| GPQA Diamond | 49.1% | 52.3% |
| MMLU-Pro | 63.1% | — |
| Confabulations | 19.1% | — |
| Vectara Hallucination Rate | — | 5.7% |
| GPQA (HELM) | 42.6% | — |
| LMArena Expert | 1245 | — |
| ARC (AI2) Challenge | 94.5% | — |
| MMLU | 85.3% | — |
| TriviaQA | 71.9% | — |
Multilingual Not comparable
Qwen2.5 72B Instruct: 41.0 (#213), Qwen3-4B: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3-4B |
|---|---|---|
| LMArena Non-English | 1252 | — |
| LMArena Chinese | 1272 | — |
| LMArena French | 1280 | — |
| LMArena German | 1234 | — |
| LMArena Japanese | 1180 | — |
| LMArena Korean | 1188 | — |
| LMArena Russian | 1264 | — |
| LMArena Spanish | 1256 | — |
Instruction Following Not comparable
Qwen2.5 72B Instruct: 65.5 (#221), Qwen3-4B: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3-4B |
|---|---|---|
| IFEval | 80.6% | — |
| LMArena Instruction Following | 1254 | — |
Long Context Not comparable
Qwen2.5 72B Instruct: 38.9 (#188), Qwen3-4B: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3-4B |
|---|---|---|
| LMArena Longer Query | 1282 | — |
Writing & Preference Not comparable
Qwen2.5 72B Instruct: 46.7 (#215), Qwen3-4B: —
| Benchmark | Qwen2.5 72B Instruct | Qwen3-4B |
|---|---|---|
| LMArena Text | 1269 | — |
| LMArena Creative Writing | 1221 | — |
| WildBench | 80.2% | — |
| LMArena Multi-Turn | 1272 | — |
Frequently asked questions
Is Qwen2.5 72B Instruct better than Qwen3-4B?
Qwen2.5 72B Instruct and Qwen3-4B score almost the same on the Noometry Index (31.9 vs 31.9), so choose on price, context window or the category you care about most.
How many benchmarks do Qwen2.5 72B Instruct and Qwen3-4B share?
2 benchmarks have published results for both models. Qwen2.5 72B Instruct has 43 scored results on Noometry and Qwen3-4B has 6.