Model comparison
Qwen2.5 7B Instruct vs Qwen3 14B
Qwen3 14B is the stronger model overall, scoring 35.0 to 28.6 on the Noometry Index. Qwen2.5 7B Instruct costs 2.0× less per token, which makes it the better buy when Qwen3 14B's lead doesn't matter for your workload.
Last verified . 6 shared benchmarks.
Alibaba (Qwen)Summary
- They share 6 benchmarks with published results for both. Qwen2.5 7B Instruct scores higher in 0 categories and Qwen3 14B in 5 categories; 4 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3 14B leads 37.0 to 11.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 2.5% for Qwen2.5 7B Instruct and 66.4% for Qwen3 14B.
- Qwen2.5 7B Instruct is cheaper at $0.17 / $0.70 per million input/output tokens, against $0.35 / $1.40 for Qwen3 14B.
Side by side
| Qwen2.5 7B Instruct | Qwen3 14B | |
|---|---|---|
| Provider | Alibaba (Qwen) | Alibaba (Qwen) |
| Noometry Index | 28.6 | 35.0 |
| Released | 2024-09 | 2025-04 |
| Weights | Open | Open |
| Context window | 131K | 131K |
| Max output | 8K | 8K |
| Input $ / M tokens | $0.17 | $0.35 |
| Output $ / M tokens | $0.70 | $1.40 |
| Results tracked | 15 | 12 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Qwen2.5 7B Instruct: 35.9 (#204), Qwen3 14B: 36.7 (#191)
| Benchmark | Qwen2.5 7B Instruct | Qwen3 14B |
|---|---|---|
| SciCode | — | 31.6% |
| BigCodeBench Instruct | 37.6% | — |
| BigCodeBench Complete | 46.1% | — |
Agentic & Tool Use Qwen3 14B leads
Qwen2.5 7B Instruct: 23.5 (#117), Qwen3 14B: 29.4 (#78)
| Benchmark | Qwen2.5 7B Instruct | Qwen3 14B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 41% |
| BALROG | 7.8% | — |
Reasoning Qwen3 14B leads
Qwen2.5 7B Instruct: 14.3 (#317), Qwen3 14B: 18.1 (#272)
| Benchmark | Qwen2.5 7B Instruct | Qwen3 14B |
|---|---|---|
| Chess Puzzles | 0% | 4% |
| DTBench | 47.7% | 64% |
| LMCA | 6.4% | 18.2% |
| Epoch Capabilities Index | 118.44 | 138.23 |
| Kagi LLM Benchmark | — | 49.1% |
| CritPt | — | 0% |
Math Qwen3 14B leads
Qwen2.5 7B Instruct: 11.8 (#300), Qwen3 14B: 37.0 (#102)
| Benchmark | Qwen2.5 7B Instruct | Qwen3 14B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 2.5% | 66.4% |
| Omni-MATH | 29.4% | — |
Knowledge Qwen3 14B leads
Qwen2.5 7B Instruct: 17.0 (#281), Qwen3 14B: 39.2 (#124)
| Benchmark | Qwen2.5 7B Instruct | Qwen3 14B |
|---|---|---|
| GPQA Diamond | 35.5% | 63.8% |
| MMLU-Pro | 53.9% | — |
| Vectara Hallucination Rate | — | 5.4% |
| GPQA (HELM) | 34.1% | — |
| MMLU | 72.9% | — |
Instruction Following Not comparable
Qwen2.5 7B Instruct: 63.4 (#226), Qwen3 14B: —
| Benchmark | Qwen2.5 7B Instruct | Qwen3 14B |
|---|---|---|
| IFEval | 74.1% | — |
Long Context Not comparable
Qwen2.5 7B Instruct: —, Qwen3 14B: 38.0 (#201)
| Benchmark | Qwen2.5 7B Instruct | Qwen3 14B |
|---|---|---|
| Fiction.LiveBench | — | 62.5% |
Writing & Preference Not comparable
Qwen2.5 7B Instruct: 48.8 (#194), Qwen3 14B: —
| Benchmark | Qwen2.5 7B Instruct | Qwen3 14B |
|---|---|---|
| WildBench | 73.1% | — |
Frequently asked questions
Is Qwen2.5 7B Instruct better than Qwen3 14B?
Qwen3 14B is the stronger model overall, scoring 35.0 to 28.6 on the Noometry Index. Qwen2.5 7B Instruct costs 2.0× less per token, which makes it the better buy when Qwen3 14B's lead doesn't matter for your workload.
Which is cheaper, Qwen2.5 7B Instruct or Qwen3 14B?
Qwen2.5 7B Instruct is cheaper. It lists at $0.17 per million input tokens and $0.70 per million output tokens; Qwen3 14B lists at $0.35 and $1.40.
Is Qwen2.5 7B Instruct or Qwen3 14B better for coding?
They score almost the same on coding (35.9 vs 36.7); test both on your own repository before choosing.
Which has the bigger context window?
Both accept 131K tokens.
How many benchmarks do Qwen2.5 7B Instruct and Qwen3 14B share?
6 benchmarks have published results for both models. Qwen2.5 7B Instruct has 15 scored results on Noometry and Qwen3 14B has 12.