Model comparison
GLM-5.3 vs Qwen3 235B-A22B
GLM-5.3 is the stronger model overall, scoring 54.8 to 43.5 on the Noometry Index. Qwen3 235B-A22B costs 1.8× less per token, which makes it the better buy when GLM-5.3's lead doesn't matter for your workload.
Last verified . 30 shared benchmarks.
Alibaba (Qwen)Summary
- They share 30 benchmarks with published results for both. GLM-5.3 scores higher in 8 categories and Qwen3 235B-A22B in 1 category; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GLM-5.3 leads 46.1 to 15.7.
- The biggest single-benchmark swing is WeirdML: 75.4% for GLM-5.3 and 41% for Qwen3 235B-A22B.
- Qwen3 235B-A22B is cheaper at $0.70 / $2.80 per million input/output tokens, against $1.40 / $4.40 for GLM-5.3.
- GLM-5.3 accepts more context: 1M tokens versus 131K.
Side by side
| GLM-5.3 | Qwen3 235B-A22B | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Alibaba (Qwen) |
| Noometry Index | 54.8 | 43.5 |
| Released | 2026-08-14 | 2025-04 |
| Weights | Open | Open |
| Context window | 1M | 131K |
| Max output | 131K | 16K |
| Input $ / M tokens | $1.40 | $0.70 |
| Output $ / M tokens | $4.40 | $2.80 |
| Results tracked | 42 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.3 leads
GLM-5.3: 59.5 (#14), Qwen3 235B-A22B: 44.3 (#75)
| Benchmark | GLM-5.3 | Qwen3 235B-A22B |
|---|---|---|
| SciCode | 59% | 42.4% |
| WeirdML | 75.4% | 41% |
| LMArena Coding | 1496 | 1445 |
| DeepSWE | 69% | — |
| FrontierCode | 40.1% | — |
| Aider Polyglot | — | 59.6% |
| CursorBench | 42.6% | — |
| LMArena WebDev | 1622 | — |
| FrontierSWE | 30.2% | — |
| ALE-Bench | 1,317 | — |
Agentic & Tool Use GLM-5.3 leads
GLM-5.3: 36.4 (#38), Qwen3 235B-A22B: 33.9 (#51)
| Benchmark | GLM-5.3 | Qwen3 235B-A22B |
|---|---|---|
| Vending-Bench 2 | 8,164 | -11.34 |
| APEX-Agents | 56.6% | — |
| Berkeley Function Calling Leaderboard | — | 52.1% |
Reasoning GLM-5.3 leads
GLM-5.3: 46.1 (#46), Qwen3 235B-A22B: 15.7 (#311)
| Benchmark | GLM-5.3 | Qwen3 235B-A22B |
|---|---|---|
| CritPt | 19.1% | 0% |
| Chess Puzzles | 21% | 12% |
| LMArena Hard Prompts | 1489 | 1433 |
| Mystery Game Puzzles | 33% | 9% |
| DTBench | 87.7% | 80.3% |
| LMCA | 55.5% | 29.3% |
| Epoch Capabilities Index | 155.61 | 143.85 |
| ARC-AGI-2 | — | 1.3% |
| SimpleBench | — | 31% |
| Kagi LLM Benchmark | — | 69.4% |
| NYT Connections (extended) | 74.2% | — |
| ARC-AGI-1 | — | 11% |
| Bench to the Future 3 | 0.15 | — |
| ForecastBench | — | 59.7 |
Math GLM-5.3 leads
GLM-5.3: 62.3 (#33), Qwen3 235B-A22B: 50.4 (#57)
| Benchmark | GLM-5.3 | Qwen3 235B-A22B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 91.1% | 86.7% |
| LMArena Math | 1489 | 1432 |
| FrontierMath (Tiers 1-3) | 68.8% | — |
| FrontierMath Tier 4 | 29.3% | — |
| ProofBench | 49% | — |
| Omni-MATH | — | 71.8% |
| MATH Level 5 | — | 68.9% |
| FrontierMath (Feb 2025 set) | — | 8.5% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge GLM-5.3 leads
GLM-5.3: 58.3 (#37), Qwen3 235B-A22B: 49.6 (#73)
| Benchmark | GLM-5.3 | Qwen3 235B-A22B |
|---|---|---|
| GPQA Diamond | 90.9% | 80.1% |
| SimpleQA Verified | 41% | 40.4% |
| LMArena Expert | 1516 | 1463 |
| MMLU-Pro | — | 84.4% |
| Confabulations | — | 15.6% |
| Vectara Hallucination Rate | — | 9.3% |
| GPQA (HELM) | — | 72.7% |
Multilingual GLM-5.3 leads
GLM-5.3: 55.7 (#28), Qwen3 235B-A22B: 52.3 (#89)
| Benchmark | GLM-5.3 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Non-English | 1457 | 1409 |
| LMArena Chinese | 1528 | 1481 |
| LMArena French | 1499 | 1445 |
| LMArena German | 1499 | 1433 |
| LMArena Japanese | 1453 | 1399 |
| LMArena Korean | 1472 | 1391 |
| LMArena Russian | 1463 | 1411 |
| LMArena Spanish | 1460 | 1430 |
Instruction Following GLM-5.3 leads
GLM-5.3: 77.5 (#23), Qwen3 235B-A22B: 72.6 (#136)
| Benchmark | GLM-5.3 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Instruction Following | 1477 | 1408 |
| IFEval | — | 83.5% |
Long Context Too close to call
GLM-5.3: 45.4 (#41), Qwen3 235B-A22B: 46.1 (#26)
| Benchmark | GLM-5.3 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Longer Query | 1482 | 1426 |
| Fiction.LiveBench | — | 75% |
Writing & Preference GLM-5.3 leads
GLM-5.3: 75.7 (#6), Qwen3 235B-A22B: 59.6 (#108)
| Benchmark | GLM-5.3 | Qwen3 235B-A22B |
|---|---|---|
| LMArena Text | 1471 | 1419 |
| LMArena Creative Writing | 1457 | 1384 |
| EQ-Bench Creative Writing | 2075 | 1366 |
| LMArena Multi-Turn | 1472 | 1432 |
| Short-Story Creative Writing | — | 83% |
| WildBench | — | 86.6% |
Frequently asked questions
Is GLM-5.3 better than Qwen3 235B-A22B?
GLM-5.3 is the stronger model overall, scoring 54.8 to 43.5 on the Noometry Index. Qwen3 235B-A22B costs 1.8× less per token, which makes it the better buy when GLM-5.3's lead doesn't matter for your workload.
Which is cheaper, GLM-5.3 or Qwen3 235B-A22B?
Qwen3 235B-A22B is cheaper. It lists at $0.70 per million input tokens and $2.80 per million output tokens; GLM-5.3 lists at $1.40 and $4.40.
Is GLM-5.3 or Qwen3 235B-A22B better for coding?
GLM-5.3 scores higher on coding benchmarks: 59.5 versus 44.3 in the Noometry coding category.
Which has the bigger context window?
GLM-5.3 does, with 1M tokens against 131K.
How many benchmarks do GLM-5.3 and Qwen3 235B-A22B share?
30 benchmarks have published results for both models. GLM-5.3 has 42 scored results on Noometry and Qwen3 235B-A22B has 49.