Model comparison
Claude Opus 5 vs GLM-5.2
Claude Opus 5 is the stronger model overall, scoring 67.8 to 51.1 on the Noometry Index. GLM-5.2 costs 4.7× less per token, which makes it the better buy when Claude Opus 5's lead doesn't matter for your workload.
Last verified . 47 shared benchmarks.
Summary
- They share 47 benchmarks with published results for both. Claude Opus 5 scores higher in 9 categories and GLM-5.2 in 0 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Claude Opus 5 leads 77.2 to 42.3.
- The biggest single-benchmark swing is GBAEval: 79.6% for Claude Opus 5 and 0% for GLM-5.2.
- GLM-5.2 is cheaper at $1.40 / $4.40 per million input/output tokens, against $5 / $25 for Claude Opus 5.
- GLM-5.2 has downloadable open weights; the other is API-only.
Side by side
| Claude Opus 5 | GLM-5.2 | |
|---|---|---|
| Provider | Anthropic | Z.ai (Zhipu) |
| Noometry Index | 67.8 | 51.1 |
| Released | 2026-07-24 | 2026-06-13 |
| Weights | Proprietary | Open |
| Context window | 1M | 1M |
| Max output | 128K | 131K |
| Input $ / M tokens | $5 | $1.40 |
| Output $ / M tokens | $25 | $4.40 |
| Results tracked | 57 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Claude Opus 5 leads
Claude Opus 5: 67.5 (#5), GLM-5.2: 51.3 (#41)
| Benchmark | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| DeepSWE | 73.6% | 43.8% |
| FrontierCode | 53.4% | 24.5% |
| LMArena WebDev | 1691 | 1603 |
| SciCode | 56.4% | 50.5% |
| WeirdML | 91.8% | 70.1% |
| LMArena Coding | 1534 | 1485 |
| ALE-Bench | 2,165 | 1,047 |
| SWE-bench Verified | — | 78.7% |
| CursorBench | 46.6% | — |
| FrontierSWE | 52% | — |
Agentic & Tool Use Claude Opus 5 leads
Claude Opus 5: 55.6 (#1), GLM-5.2: 32.4 (#63)
| Benchmark | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| APEX-Agents | 65.8% | 45.2% |
| τ²-bench Banking | 48.7% | 37.1% |
| PostTrainBench | 35% | 31.7% |
| GBAEval | 79.6% | 0% |
| Vending-Bench 2 | 11,182 | 8,314 |
| OSWorld 2.0 | 31.4% | — |
| BALROG | 63.4% | — |
| GDP.pdf | 24% | — |
Reasoning Claude Opus 5 leads
Claude Opus 5: 77.2 (#4), GLM-5.2: 42.3 (#52)
| Benchmark | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| ARC-AGI-2 | 90.4% | 22.8% |
| SimpleBench | 80.6% | 58.8% |
| NYT Connections (extended) | 94.3% | 74.3% |
| ARC-AGI-1 | 97.5% | 77% |
| CritPt | 29.1% | 20.9% |
| Chess Puzzles | 42% | 21% |
| EBR-Bench | 45.7% | 9.5% |
| LMArena Hard Prompts | 1526 | 1480 |
| Mystery Game Puzzles | 59% | 19% |
| DTBench | 97.9% | 93.6% |
| LMCA | 64.5% | 45.8% |
| Epoch Capabilities Index | 162.78 | 151.78 |
| Kagi LLM Benchmark | — | 62.6% |
| Surface Evolver Bench | — | 55.6% |
| Bench to the Future 3 | 0.12 | — |
Math Claude Opus 5 leads
Claude Opus 5: 86.2 (#8), GLM-5.2: 55.7 (#43)
| Benchmark | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 85.6% | 59.2% |
| FrontierMath Tier 4 | 73.2% | 29.3% |
| OTIS Mock AIME 2024-2025 | 98.9% | 86.4% |
| ProofBench | 99% | 35% |
| LMArena Math | 1531 | 1482 |
| MathArena Final-Answer Competitions | — | 67.6% |
Knowledge Claude Opus 5 leads
Claude Opus 5: 66.8 (#9), GLM-5.2: 57.1 (#40)
| Benchmark | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| GPQA Diamond | 93.9% | 91.9% |
| SimpleQA Verified | 59.9% | 34.2% |
| LMArena Expert | 1557 | 1486 |
Multimodal Not comparable
Claude Opus 5: 50.8 (#8), GLM-5.2: —
| Benchmark | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| LMArena Vision | 1319 | — |
| Blueprint-Bench 2 | 30.4% | — |
| Furniture Assembly | 60.8% | — |
| LMArena Document | 1516 | — |
Multilingual Claude Opus 5 leads
Claude Opus 5: 58.8 (#4), GLM-5.2: 55.8 (#26)
| Benchmark | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| LMArena Non-English | 1501 | 1459 |
| LMArena Chinese | 1574 | 1519 |
| LMArena French | 1519 | 1479 |
| LMArena German | 1524 | 1468 |
| LMArena Japanese | 1516 | 1451 |
| LMArena Korean | 1521 | 1445 |
| LMArena Russian | 1507 | 1466 |
| LMArena Spanish | 1519 | 1477 |
Instruction Following Claude Opus 5 leads
Claude Opus 5: 79.2 (#7), GLM-5.2: 76.9 (#34)
| Benchmark | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| LMArena Instruction Following | 1517 | 1465 |
Long Context Claude Opus 5 leads
Claude Opus 5: 46.5 (#21), GLM-5.2: 45.3 (#43)
| Benchmark | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| LMArena Longer Query | 1515 | 1479 |
Writing & Preference Claude Opus 5 leads
Claude Opus 5: 79.2 (#1), GLM-5.2: 70.4 (#21)
| Benchmark | Claude Opus 5 | GLM-5.2 |
|---|---|---|
| LMArena Text | 1507 | 1470 |
| LMArena Creative Writing | 1491 | 1462 |
| EQ-Bench Creative Writing | 2133 | 1757 |
| EQ-Bench 4 | 1385 | 1222 |
| LMArena Multi-Turn | 1499 | 1469 |
Frequently asked questions
Is Claude Opus 5 better than GLM-5.2?
Claude Opus 5 is the stronger model overall, scoring 67.8 to 51.1 on the Noometry Index. GLM-5.2 costs 4.7× less per token, which makes it the better buy when Claude Opus 5's lead doesn't matter for your workload.
Which is cheaper, Claude Opus 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 Opus 5 lists at $5 and $25.
Is Claude Opus 5 or GLM-5.2 better for coding?
Claude Opus 5 scores higher on coding benchmarks: 67.5 versus 51.3 in the Noometry coding category.
Which has the bigger context window?
Both accept 1M tokens.
How many benchmarks do Claude Opus 5 and GLM-5.2 share?
47 benchmarks have published results for both models. Claude Opus 5 has 57 scored results on Noometry and GLM-5.2 has 51.