Model comparison
Qwen2.5 7B Instruct vs Qwen3-4B
Qwen3-4B is the stronger model overall, scoring 31.9 to 29.0 on the Noometry Index.
Last verified . 3 shared benchmarks.
Alibaba (Qwen)Summary
- They share 3 benchmarks with published results for both. Qwen2.5 7B Instruct scores higher in 0 categories and Qwen3-4B in 4 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3-4B leads 29.7 to 12.6.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 2.5% for Qwen2.5 7B Instruct and 52.2% for Qwen3-4B.
Side by side
| Qwen2.5 7B Instruct | Qwen3-4B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 29.0 | 31.9 |
| Released | 2024-09 | 2025-04-29 |
| Weights | Open | Open |
| Context window | 131K | — |
| Max output | 8K | — |
| Input $ / M tokens | $0.17 | — |
| Output $ / M tokens | $0.70 | — |
| Results tracked | 15 | 6 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Not comparable
Qwen2.5 7B Instruct: 36.5 (#208), Qwen3-4B: —
| Benchmark | Qwen2.5 7B Instruct | Qwen3-4B |
|---|---|---|
| BigCodeBench Instruct | 37.6% | — |
| BigCodeBench Complete | 46.1% | — |
Agentic & Tool Use Qwen3-4B leads
Qwen2.5 7B Instruct: 23.8 (#124), Qwen3-4B: 27.6 (#100)
| Benchmark | Qwen2.5 7B Instruct | Qwen3-4B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 35.7% |
| BALROG | 7.8% | — |
Reasoning Qwen3-4B leads
Qwen2.5 7B Instruct: 14.8 (#322), Qwen3-4B: 19.2 (#268)
| Benchmark | Qwen2.5 7B Instruct | Qwen3-4B |
|---|---|---|
| Chess Puzzles | 0% | 4% |
| DTBench | 47.7% | — |
| LMCA | 6.4% | — |
| Epoch Capabilities Index | 118.51 | — |
Math Qwen3-4B leads
Qwen2.5 7B Instruct: 12.6 (#306), Qwen3-4B: 29.7 (#240)
| Benchmark | Qwen2.5 7B Instruct | Qwen3-4B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 2.5% | 52.2% |
| MathArena Final-Answer Competitions | — | 38.5% |
| Omni-MATH | 29.4% | — |
Knowledge Qwen3-4B leads
Qwen2.5 7B Instruct: 17.0 (#286), Qwen3-4B: 33.0 (#208)
| Benchmark | Qwen2.5 7B Instruct | Qwen3-4B |
|---|---|---|
| GPQA Diamond | 35.5% | 52.3% |
| MMLU-Pro | 53.9% | — |
| Vectara Hallucination Rate | — | 5.7% |
| GPQA (HELM) | 34.1% | — |
| MMLU | 72.9% | — |
Instruction Following Not comparable
Qwen2.5 7B Instruct: 63.2 (#231), Qwen3-4B: —
| Benchmark | Qwen2.5 7B Instruct | Qwen3-4B |
|---|---|---|
| IFEval | 74.1% | — |
Writing & Preference Not comparable
Qwen2.5 7B Instruct: 48.8 (#195), Qwen3-4B: —
| Benchmark | Qwen2.5 7B Instruct | Qwen3-4B |
|---|---|---|
| WildBench | 73.1% | — |
Frequently asked questions
Is Qwen2.5 7B Instruct better than Qwen3-4B?
Qwen3-4B is the stronger model overall, scoring 31.9 to 29.0 on the Noometry Index.
How many benchmarks do Qwen2.5 7B Instruct and Qwen3-4B share?
3 benchmarks have published results for both models. Qwen2.5 7B Instruct has 15 scored results on Noometry and Qwen3-4B has 6.