Model comparison
Claude Sonnet 4.5 vs GLM-5.2
GLM-5.2 is the stronger model overall, scoring 51.1 to 44.1 on the Noometry Index.
Last verified . 43 shared benchmarks.
Summary
- They share 43 benchmarks with published results for both. Claude Sonnet 4.5 scores higher in 1 category and GLM-5.2 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-5.2 leads 55.7 to 32.3.
- The biggest single-benchmark swing is NYT Connections (extended): 37.3% for Claude Sonnet 4.5 and 74.3% for GLM-5.2.
- GLM-5.2 is cheaper at $1.40 / $4.40 per million input/output tokens, against $3 / $15 for Claude Sonnet 4.5.
- GLM-5.2 accepts more context: 1M tokens versus 200K.
- GLM-5.2 has downloadable open weights; the other is API-only.
Side by side
| Claude Sonnet 4.5 | GLM-5.2 | |
|---|---|---|
| Provider | Anthropic | Z.ai (Zhipu) |
| Noometry Index | 44.1 | 51.1 |
| Released | 2025-09-29 | 2026-06-13 |
| Weights | Proprietary | Open |
| Context window | 200K | 1M |
| Max output | 64K | 131K |
| Input $ / M tokens | $3 | $1.40 |
| Output $ / M tokens | $15 | $4.40 |
| Results tracked | 73 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.2 leads
Claude Sonnet 4.5: 47.3 (#61), GLM-5.2: 51.3 (#41)
| Benchmark | Claude Sonnet 4.5 | GLM-5.2 |
|---|---|---|
| SWE-bench Verified | 71.3% | 78.7% |
| LMArena WebDev | 1393 | 1603 |
| SciCode | 44.7% | 50.5% |
| WeirdML | 47.7% | 70.1% |
| LMArena Coding | 1489 | 1485 |
| ALE-Bench | 796.15 | 1,047 |
| DeepSWE | — | 43.8% |
| FrontierCode | — | 24.5% |
| SWE-bench Verified (bash only) | 71.4% | — |
| SWE-bench Multilingual | 67% | — |
| GSO | 14.7% | — |
| AlgoTune | 1.52 | — |
Agentic & Tool Use Claude Sonnet 4.5 leads
Claude Sonnet 4.5: 38.3 (#32), GLM-5.2: 32.4 (#63)
| Benchmark | Claude Sonnet 4.5 | GLM-5.2 |
|---|---|---|
| τ²-bench Banking | 25.3% | 37.1% |
| Vending-Bench 2 | 3,839 | 8,314 |
| Terminal-Bench | 46.5% | — |
| APEX-Agents | — | 45.2% |
| Berkeley Function Calling Leaderboard | 73.2% | — |
| GDPval | 42.5% | — |
| Remote Labor Index | 2.1% | — |
| τ²-bench Airline | 72% | — |
| τ²-bench Retail | 72.4% | — |
| τ²-bench Telecom | 84.9% | — |
| Cybench | 60% | — |
| DeepResearch Bench | 52.6% | — |
| OSWorld | 62.9% | — |
| PostTrainBench | — | 31.7% |
| GBAEval | — | 0% |
| LMArena Search | 1159 | — |
| METR Time Horizons | 67.4% | — |
Reasoning GLM-5.2 leads
Claude Sonnet 4.5: 26.9 (#125), GLM-5.2: 42.3 (#52)
| Benchmark | Claude Sonnet 4.5 | GLM-5.2 |
|---|---|---|
| ARC-AGI-2 | 13.6% | 22.8% |
| SimpleBench | 54.3% | 58.8% |
| Kagi LLM Benchmark | 57.9% | 62.6% |
| NYT Connections (extended) | 37.3% | 74.3% |
| ARC-AGI-1 | 63.7% | 77% |
| CritPt | 1.1% | 20.9% |
| Chess Puzzles | 12% | 21% |
| EBR-Bench | 2.4% | 9.5% |
| LMArena Hard Prompts | 1462 | 1480 |
| Mystery Game Puzzles | 17% | 19% |
| DTBench | 83.2% | 93.6% |
| LMCA | 38.8% | 45.8% |
| Epoch Capabilities Index | 146.84 | 151.78 |
| EnigmaEval | 6% | — |
| Surface Evolver Bench | — | 55.6% |
| ForecastBench | 61.9 | — |
Math GLM-5.2 leads
Claude Sonnet 4.5: 32.3 (#216), GLM-5.2: 55.7 (#43)
| Benchmark | Claude Sonnet 4.5 | GLM-5.2 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 23.9% | 59.2% |
| FrontierMath Tier 4 | 2.4% | 29.3% |
| OTIS Mock AIME 2024-2025 | 77.8% | 86.4% |
| ProofBench | 19% | 35% |
| LMArena Math | 1449 | 1482 |
| MathArena Final-Answer Competitions | — | 67.6% |
| Omni-MATH | 55.3% | — |
| MATH Level 5 | 97.7% | — |
| FrontierMath (Feb 2025 set) | 15.2% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |
Knowledge GLM-5.2 leads
Claude Sonnet 4.5: 48.4 (#76), GLM-5.2: 57.1 (#40)
| Benchmark | Claude Sonnet 4.5 | GLM-5.2 |
|---|---|---|
| GPQA Diamond | 82.3% | 91.9% |
| SimpleQA Verified | 30.7% | 34.2% |
| LMArena Expert | 1482 | 1486 |
| Humanity's Last Exam | 13.7% | — |
| MMLU-Pro | 86.9% | — |
| Vectara Hallucination Rate | 12% | — |
| GPQA (HELM) | 68.6% | — |
Multimodal Not comparable
Claude Sonnet 4.5: 34.8 (#89), GLM-5.2: —
| Benchmark | Claude Sonnet 4.5 | GLM-5.2 |
|---|---|---|
| VPCT | 39.8% | — |
| LMArena Document | 1450 | — |
Multilingual GLM-5.2 leads
Claude Sonnet 4.5: 53.4 (#69), GLM-5.2: 55.8 (#26)
| Benchmark | Claude Sonnet 4.5 | GLM-5.2 |
|---|---|---|
| LMArena Non-English | 1425 | 1459 |
| LMArena Chinese | 1459 | 1519 |
| LMArena French | 1458 | 1479 |
| LMArena German | 1427 | 1468 |
| LMArena Japanese | 1390 | 1451 |
| LMArena Korean | 1403 | 1445 |
| LMArena Russian | 1437 | 1466 |
| LMArena Spanish | 1457 | 1477 |
Instruction Following GLM-5.2 leads
Claude Sonnet 4.5: 75.0 (#78), GLM-5.2: 76.9 (#34)
| Benchmark | Claude Sonnet 4.5 | GLM-5.2 |
|---|---|---|
| LMArena Instruction Following | 1459 | 1465 |
| IFEval | 85% | — |
Long Context Too close to call
Claude Sonnet 4.5: 45.2 (#46), GLM-5.2: 45.3 (#43)
| Benchmark | Claude Sonnet 4.5 | GLM-5.2 |
|---|---|---|
| LMArena Longer Query | 1476 | 1479 |
Writing & Preference GLM-5.2 leads
Claude Sonnet 4.5: 66.5 (#34), GLM-5.2: 70.4 (#21)
| Benchmark | Claude Sonnet 4.5 | GLM-5.2 |
|---|---|---|
| LMArena Text | 1439 | 1470 |
| LMArena Creative Writing | 1442 | 1462 |
| EQ-Bench Creative Writing | 1678 | 1757 |
| LMArena Multi-Turn | 1465 | 1469 |
| WildBench | 85.4% | — |
| EQ-Bench 4 | — | 1222 |
Frequently asked questions
Is Claude Sonnet 4.5 better than GLM-5.2?
GLM-5.2 is the stronger model overall, scoring 51.1 to 44.1 on the Noometry Index.
Which is cheaper, Claude Sonnet 4.5 or GLM-5.2?
GLM-5.2 is cheaper. It lists at $1.40 per million input tokens and $4.40 per million output tokens; Claude Sonnet 4.5 lists at $3 and $15.
Is Claude Sonnet 4.5 or GLM-5.2 better for coding?
GLM-5.2 scores higher on coding benchmarks: 51.3 versus 47.3 in the Noometry coding category.
Which has the bigger context window?
GLM-5.2 does, with 1M tokens against 200K.
How many benchmarks do Claude Sonnet 4.5 and GLM-5.2 share?
43 benchmarks have published results for both models. Claude Sonnet 4.5 has 73 scored results on Noometry and GLM-5.2 has 51.