Model comparison
GLM-5.3-Flash vs Qwen2.5 7B Instruct
GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 29.0 on the Noometry Index.
Last verified . 4 shared benchmarks.
Alibaba (Qwen)Summary
- They share 4 benchmarks with published results for both. GLM-5.3-Flash scores higher in 7 categories and Qwen2.5 7B Instruct in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-5.3-Flash leads 58.4 to 17.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 93.9% for GLM-5.3-Flash and 2.5% for Qwen2.5 7B Instruct.
- GLM-5.3-Flash is cheaper at $0.15 / $0.50 per million input/output tokens, against $0.17 / $0.70 for Qwen2.5 7B Instruct.
- GLM-5.3-Flash accepts more context: 1M tokens versus 131K.
Side by side
| GLM-5.3-Flash | Qwen2.5 7B Instruct | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Alibaba (Qwen) |
| Noometry Index | 51.8 | 29.0 |
| Released | 2026-08-20 | 2024-09 |
| Weights | Open | Open |
| Context window | 1M | 131K |
| Max output | 131K | 8K |
| Input $ / M tokens | $0.15 | $0.17 |
| Output $ / M tokens | $0.50 | $0.70 |
| Results tracked | 40 | 15 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3-Flash leads
GLM-5.3-Flash: 53.1 (#31), Qwen2.5 7B Instruct: 36.5 (#208)
| Benchmark | GLM-5.3-Flash | Qwen2.5 7B Instruct |
|---|---|---|
| DeepSWE | 63.4% | — |
| FrontierCode | 31.8% | — |
| CursorBench | 36.8% | — |
| LMArena WebDev | 1609 | — |
| FrontierSWE | 18.1% | — |
| SciCode | 51.6% | — |
| BigCodeBench Instruct | — | 37.6% |
| LMArena Coding | 1508 | — |
| BigCodeBench Complete | — | 46.1% |
| ALE-Bench | 303.55 | — |
Agentic & Tool Use GLM-5.3-Flash leads
GLM-5.3-Flash: 34.2 (#47), Qwen2.5 7B Instruct: 23.8 (#124)
| Benchmark | GLM-5.3-Flash | Qwen2.5 7B Instruct |
|---|---|---|
| APEX-Agents | 52.8% | — |
| BALROG | — | 7.8% |
| GDP.pdf | 14% | — |
Reasoning GLM-5.3-Flash leads
GLM-5.3-Flash: 48.0 (#42), Qwen2.5 7B Instruct: 14.8 (#322)
| Benchmark | GLM-5.3-Flash | Qwen2.5 7B Instruct |
|---|---|---|
| Chess Puzzles | 14% | 0% |
| Epoch Capabilities Index | 151.88 | 118.51 |
| ARC-AGI-2 | 65.8% | — |
| ARC-AGI-1 | 91% | — |
| CritPt | 15.4% | — |
| LMArena Hard Prompts | 1491 | — |
| Mystery Game Puzzles | 8% | — |
| DTBench | — | 47.7% |
| LMCA | — | 6.4% |
| Surface Evolver Bench | 52.5% | — |
| Bench to the Future 3 | 0.15 | — |
Math GLM-5.3-Flash leads
GLM-5.3-Flash: 53.3 (#47), Qwen2.5 7B Instruct: 12.6 (#306)
| Benchmark | GLM-5.3-Flash | Qwen2.5 7B Instruct |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 93.9% | 2.5% |
| FrontierMath (Tiers 1-3) | 55.8% | — |
| FrontierMath Tier 4 | 17.1% | — |
| ProofBench | 21% | — |
| Omni-MATH | — | 29.4% |
| LMArena Math | 1500 | — |
Knowledge GLM-5.3-Flash leads
GLM-5.3-Flash: 58.4 (#36), Qwen2.5 7B Instruct: 17.0 (#286)
| Benchmark | GLM-5.3-Flash | Qwen2.5 7B Instruct |
|---|---|---|
| GPQA Diamond | 90.2% | 35.5% |
| MMLU-Pro | — | 53.9% |
| GPQA (HELM) | — | 34.1% |
| LMArena Expert | 1513 | — |
| MMLU | — | 72.9% |
Multimodal Not comparable
GLM-5.3-Flash: 42.8 (#27), Qwen2.5 7B Instruct: —
| Benchmark | GLM-5.3-Flash | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Vision | 1296 | — |
Multilingual Not comparable
GLM-5.3-Flash: 56.0 (#25), Qwen2.5 7B Instruct: —
| Benchmark | GLM-5.3-Flash | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Non-English | 1462 | — |
| LMArena Chinese | 1527 | — |
| LMArena French | 1496 | — |
| LMArena German | 1470 | — |
| LMArena Japanese | 1429 | — |
| LMArena Korean | 1446 | — |
| LMArena Russian | 1469 | — |
| LMArena Spanish | 1471 | — |
Instruction Following GLM-5.3-Flash leads
GLM-5.3-Flash: 77.5 (#20), Qwen2.5 7B Instruct: 63.2 (#231)
| Benchmark | GLM-5.3-Flash | Qwen2.5 7B Instruct |
|---|---|---|
| IFEval | — | 74.1% |
| LMArena Instruction Following | 1478 | — |
Long Context Not comparable
GLM-5.3-Flash: 45.4 (#39), Qwen2.5 7B Instruct: —
| Benchmark | GLM-5.3-Flash | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Longer Query | 1482 | — |
Writing & Preference GLM-5.3-Flash leads
GLM-5.3-Flash: 65.3 (#50), Qwen2.5 7B Instruct: 48.8 (#195)
| Benchmark | GLM-5.3-Flash | Qwen2.5 7B Instruct |
|---|---|---|
| LMArena Text | 1471 | — |
| LMArena Creative Writing | 1442 | — |
| WildBench | — | 73.1% |
| LMArena Multi-Turn | 1467 | — |
Frequently asked questions
Is GLM-5.3-Flash better than Qwen2.5 7B Instruct?
GLM-5.3-Flash is the stronger model overall, scoring 51.8 to 29.0 on the Noometry Index.
Which is cheaper, GLM-5.3-Flash or Qwen2.5 7B Instruct?
GLM-5.3-Flash is cheaper. It lists at $0.15 per million input tokens and $0.50 per million output tokens; Qwen2.5 7B Instruct lists at $0.17 and $0.70.
Is GLM-5.3-Flash or Qwen2.5 7B Instruct better for coding?
GLM-5.3-Flash scores higher on coding benchmarks: 53.1 versus 36.5 in the Noometry coding category.
Which has the bigger context window?
GLM-5.3-Flash does, with 1M tokens against 131K.
How many benchmarks do GLM-5.3-Flash and Qwen2.5 7B Instruct share?
4 benchmarks have published results for both models. GLM-5.3-Flash has 40 scored results on Noometry and Qwen2.5 7B Instruct has 15.