Model comparison
Claude Sonnet 4 vs GLM-4.5-Air
Claude Sonnet 4 is the stronger model overall, scoring 40.6 to 38.7 on the Noometry Index. GLM-4.5-Air costs 14× less per token, which makes it the better buy when Claude Sonnet 4's lead doesn't matter for your workload.
Last verified . 27 shared benchmarks.
Summary
- They share 27 benchmarks with published results for both. Claude Sonnet 4 scores higher in 5 categories and GLM-4.5-Air in 3 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in coding, where Claude Sonnet 4 leads 44.0 to 34.5.
- The biggest single-benchmark swing is Kagi LLM Benchmark: 73% for Claude Sonnet 4 and 43% for GLM-4.5-Air.
- GLM-4.5-Air is cheaper at $0.20 / $1.10 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.
- Claude Sonnet 4 accepts more context: 200K tokens versus 131K.
- GLM-4.5-Air has downloadable open weights; the other is API-only.
Side by side
| Claude Sonnet 4 | GLM-4.5-Air | |
|---|---|---|
| Provider | Anthropic | Z.ai (Zhipu) |
| Noometry Index | 40.6 | 38.7 |
| Released | 2025-05-22 | 2025-07-20 |
| Weights | Proprietary | Open |
| Context window | 200K | 131K |
| Max output | 64K | 98K |
| Input $ / M tokens | $3 | $0.20 |
| Output $ / M tokens | $15 | $1.10 |
| Results tracked | 58 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Sonnet 4 leads
Claude Sonnet 4: 44.0 (#70), GLM-4.5-Air: 34.5 (#225)
| Benchmark | Claude Sonnet 4 | GLM-4.5-Air |
|---|---|---|
| GSO | 4.9% | 2.9% |
| LMArena Coding | 1414 | 1397 |
| SWE-bench Verified (bash only) | 64.9% | — |
| Aider Polyglot | 61.3% | — |
| SciCode | 40% | — |
| WeirdML | 46.1% | — |
| ALE-Bench | 655.35 | — |
Agentic & Tool Use Not comparable
Claude Sonnet 4: 38.3 (#29), GLM-4.5-Air: —
| Benchmark | Claude Sonnet 4 | GLM-4.5-Air |
|---|---|---|
| TheAgentCompany | 33.1% | — |
| Cybench | 35% | — |
| DeepResearch Bench | 46.6% | — |
| OSWorld | 43.9% | — |
| METR Time Horizons | 62% | — |
Reasoning Too close to call
Claude Sonnet 4: 23.0 (#169), GLM-4.5-Air: 23.3 (#166)
| Benchmark | Claude Sonnet 4 | GLM-4.5-Air |
|---|---|---|
| Kagi LLM Benchmark | 73% | 43% |
| LMArena Hard Prompts | 1372 | 1379 |
| ForecastBench | 60.2 | 59.2 |
| ARC-AGI-2 | 5.9% | — |
| SimpleBench | 45.5% | — |
| ARC-AGI-1 | 40% | — |
| CritPt | 0.3% | — |
| EnigmaEval | 3.1% | — |
| DTBench | 77.1% | — |
| LMCA | 29% | — |
| Epoch Capabilities Index | 141.69 | — |
Math Claude Sonnet 4 leads
Claude Sonnet 4: 41.5 (#71), GLM-4.5-Air: 33.8 (#166)
| Benchmark | Claude Sonnet 4 | GLM-4.5-Air |
|---|---|---|
| Omni-MATH | 60.2% | 39.1% |
| LMArena Math | 1375 | 1396 |
| OTIS Mock AIME 2024-2025 | 71.1% | — |
| MATH Level 5 | 84.4% | — |
| FrontierMath (Feb 2025 set) | 4.1% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Claude Sonnet 4 leads
Claude Sonnet 4: 41.6 (#99), GLM-4.5-Air: 34.6 (#184)
| Benchmark | Claude Sonnet 4 | GLM-4.5-Air |
|---|---|---|
| Humanity's Last Exam | 7.8% | 8.1% |
| MMLU-Pro | 84.3% | 76.2% |
| Vectara Hallucination Rate | 10.3% | 9.3% |
| GPQA (HELM) | 70.6% | 59.4% |
| LMArena Expert | 1371 | 1370 |
| GPQA Diamond | 79.2% | — |
| Confabulations | 13.2% | — |
Multimodal Not comparable
Claude Sonnet 4: 25.6 (#116), GLM-4.5-Air: —
| Benchmark | Claude Sonnet 4 | GLM-4.5-Air |
|---|---|---|
| LMArena Vision | 1191 | — |
| GeoBench | 37% | — |
| VPCT | 34% | — |
| MindCube | 44.8% | — |
Multilingual GLM-4.5-Air leads
Claude Sonnet 4: 46.9 (#151), GLM-4.5-Air: 49.4 (#130)
| Benchmark | Claude Sonnet 4 | GLM-4.5-Air |
|---|---|---|
| LMArena Non-English | 1333 | 1366 |
| LMArena Chinese | 1351 | 1427 |
| LMArena French | 1364 | 1399 |
| LMArena German | 1331 | 1376 |
| LMArena Japanese | 1301 | 1348 |
| LMArena Korean | 1293 | 1306 |
| LMArena Russian | 1355 | 1373 |
| LMArena Spanish | 1356 | 1385 |
Instruction Following Claude Sonnet 4 leads
Claude Sonnet 4: 71.8 (#141), GLM-4.5-Air: 69.8 (#165)
| Benchmark | Claude Sonnet 4 | GLM-4.5-Air |
|---|---|---|
| IFEval | 84% | 81.2% |
| LMArena Instruction Following | 1375 | 1353 |
Long Context GLM-4.5-Air leads
Claude Sonnet 4: 33.6 (#254), GLM-4.5-Air: 41.6 (#130)
| Benchmark | Claude Sonnet 4 | GLM-4.5-Air |
|---|---|---|
| LMArena Longer Query | 1398 | 1367 |
| Fiction.LiveBench | 46.9% | — |
Writing & Preference Claude Sonnet 4 leads
Claude Sonnet 4: 57.3 (#127), GLM-4.5-Air: 56.1 (#134)
| Benchmark | Claude Sonnet 4 | GLM-4.5-Air |
|---|---|---|
| LMArena Text | 1351 | 1384 |
| LMArena Creative Writing | 1345 | 1343 |
| WildBench | 83.8% | 78.9% |
| LMArena Multi-Turn | 1376 | 1371 |
| Short-Story Creative Writing | 81.4% | — |
| EQ-Bench Creative Writing | 1483 | — |
Frequently asked questions
Is Claude Sonnet 4 better than GLM-4.5-Air?
Claude Sonnet 4 is the stronger model overall, scoring 40.6 to 38.7 on the Noometry Index. GLM-4.5-Air costs 14× less per token, which makes it the better buy when Claude Sonnet 4's lead doesn't matter for your workload.
Which is cheaper, Claude Sonnet 4 or GLM-4.5-Air?
GLM-4.5-Air is cheaper. It lists at $0.20 per million input tokens and $1.10 per million output tokens; Claude Sonnet 4 lists at $3 and $15.
Is Claude Sonnet 4 or GLM-4.5-Air better for coding?
Claude Sonnet 4 scores higher on coding benchmarks: 44.0 versus 34.5 in the Noometry coding category.
Which has the bigger context window?
Claude Sonnet 4 does, with 200K tokens against 131K.
How many benchmarks do Claude Sonnet 4 and GLM-4.5-Air share?
27 benchmarks have published results for both models. Claude Sonnet 4 has 58 scored results on Noometry and GLM-4.5-Air has 27.