Model comparison
GLM-5.2 vs o1
GLM-5.2 is the stronger model overall, scoring 51.1 to 40.9 on the Noometry Index.
Last verified . 28 shared benchmarks.
Summary
- They share 28 benchmarks with published results for both. GLM-5.2 scores higher in 8 categories and o1 in 1 category; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-5.2 leads 55.7 to 36.1.
- The biggest single-benchmark swing is ARC-AGI-1: 77% for GLM-5.2 and 30.7% for o1.
- GLM-5.2 is cheaper at $1.40 / $4.40 per million input/output tokens, against $15 / $60 for o1.
- 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
| GLM-5.2 | o1 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | OpenAI |
| Noometry Index | 51.1 | 40.9 |
| Released | 2026-06-13 | 2024-09-12 |
| Weights | Open | Proprietary |
| Context window | 1M | 200K |
| Max output | 131K | 100K |
| Input $ / M tokens | $1.40 | $15 |
| Output $ / M tokens | $4.40 | $60 |
| Results tracked | 51 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.2 leads
GLM-5.2: 51.3 (#41), o1: 46.1 (#70)
| Benchmark | GLM-5.2 | o1 |
|---|---|---|
| WeirdML | 70.1% | 47.6% |
| LMArena Coding | 1485 | 1367 |
| SWE-bench Verified | 78.7% | — |
| DeepSWE | 43.8% | — |
| FrontierCode | 24.5% | — |
| Aider Polyglot | — | 61.7% |
| LMArena WebDev | 1603 | — |
| SciCode | 50.5% | — |
| LiveBench Coding | — | 69.7% |
| CadEval | — | 56% |
| ALE-Bench | 1,047 | — |
| HumanEval+ | — | 89% |
| MBPP+ | — | 80.2% |
Agentic & Tool Use GLM-5.2 leads
GLM-5.2: 32.4 (#63), o1: 24.6 (#117)
| Benchmark | GLM-5.2 | o1 |
|---|---|---|
| APEX-Agents | 45.2% | — |
| τ²-bench Banking | 37.1% | — |
| Cybench | — | 10% |
| PostTrainBench | 31.7% | — |
| GBAEval | 0% | — |
| METR Time Horizons | — | 51.1% |
| Vending-Bench 2 | 8,314 | — |
Reasoning GLM-5.2 leads
GLM-5.2: 42.3 (#52), o1: 27.9 (#111)
| Benchmark | GLM-5.2 | o1 |
|---|---|---|
| SimpleBench | 58.8% | 41.7% |
| ARC-AGI-1 | 77% | 30.7% |
| Chess Puzzles | 21% | 15% |
| LMArena Hard Prompts | 1480 | 1371 |
| DTBench | 93.6% | 74.7% |
| LMCA | 45.8% | 22.3% |
| Epoch Capabilities Index | 151.78 | 141.91 |
| ARC-AGI-2 | 22.8% | — |
| Kagi LLM Benchmark | 62.6% | — |
| NYT Connections (extended) | 74.3% | — |
| CritPt | 20.9% | — |
| EnigmaEval | — | 5.7% |
| EBR-Bench | 9.5% | — |
| LiveBench Reasoning | — | 91.6% |
| Mystery Game Puzzles | 19% | — |
| LiveBench Data Analysis | — | 65.5% |
| Surface Evolver Bench | 55.6% | — |
| LiveBench | — | 75.7% |
Math GLM-5.2 leads
GLM-5.2: 55.7 (#43), o1: 36.1 (#175)
| Benchmark | GLM-5.2 | o1 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 59.2% | 14.7% |
| OTIS Mock AIME 2024-2025 | 86.4% | 73.3% |
| LMArena Math | 1482 | 1388 |
| FrontierMath Tier 4 | 29.3% | — |
| MathArena Final-Answer Competitions | 67.6% | — |
| ProofBench | 35% | — |
| LiveBench Math | — | 80.3% |
| MATH Level 5 | — | 94.7% |
| FrontierMath (Feb 2025 set) | — | 9.3% |
Knowledge GLM-5.2 leads
GLM-5.2: 57.1 (#40), o1: 41.5 (#110)
| Benchmark | GLM-5.2 | o1 |
|---|---|---|
| GPQA Diamond | 91.9% | 76.8% |
| SimpleQA Verified | 34.2% | 41.1% |
| LMArena Expert | 1486 | 1361 |
| Humanity's Last Exam | — | 8% |
| Confabulations | — | 11.7% |
Multimodal Not comparable
GLM-5.2: —, o1: 34.2 (#93)
| Benchmark | GLM-5.2 | o1 |
|---|---|---|
| LMArena Vision | — | 1168 |
| GeoBench | — | 80% |
| VPCT | — | 37% |
| SpatialViz-Bench | — | 41.4% |
Multilingual GLM-5.2 leads
GLM-5.2: 55.8 (#26), o1: 48.6 (#142)
| Benchmark | GLM-5.2 | o1 |
|---|---|---|
| LMArena Non-English | 1459 | 1358 |
| LMArena Chinese | 1519 | 1394 |
| LMArena French | 1479 | 1344 |
| LMArena German | 1468 | 1337 |
| LMArena Japanese | 1451 | 1346 |
| LMArena Korean | 1445 | 1396 |
| LMArena Russian | 1466 | 1356 |
| LMArena Spanish | 1477 | 1345 |
Instruction Following GLM-5.2 leads
GLM-5.2: 76.9 (#34), o1: 74.8 (#86)
| Benchmark | GLM-5.2 | o1 |
|---|---|---|
| LMArena Instruction Following | 1465 | 1367 |
| LiveBench Instruction Following | — | 81.5% |
Long Context o1 leads
GLM-5.2: 45.3 (#43), o1: 50.3 (#9)
| Benchmark | GLM-5.2 | o1 |
|---|---|---|
| LMArena Longer Query | 1479 | 1378 |
| Fiction.LiveBench | — | 83.3% |
Writing & Preference GLM-5.2 leads
GLM-5.2: 70.4 (#21), o1: 55.6 (#144)
| Benchmark | GLM-5.2 | o1 |
|---|---|---|
| LMArena Text | 1470 | 1366 |
| LMArena Creative Writing | 1462 | 1348 |
| LMArena Multi-Turn | 1469 | 1369 |
| Short-Story Creative Writing | — | 70.2% |
| EQ-Bench Creative Writing | 1757 | — |
| EQ-Bench 4 | 1222 | — |
| LiveBench Language | — | 65.4% |
Frequently asked questions
Is GLM-5.2 better than o1?
GLM-5.2 is the stronger model overall, scoring 51.1 to 40.9 on the Noometry Index.
Which is cheaper, GLM-5.2 or o1?
GLM-5.2 is cheaper. It lists at $1.40 per million input tokens and $4.40 per million output tokens; o1 lists at $15 and $60.
Is GLM-5.2 or o1 better for coding?
GLM-5.2 scores higher on coding benchmarks: 51.3 versus 46.1 in the Noometry coding category.
Which has the bigger context window?
GLM-5.2 does, with 1M tokens against 200K.
How many benchmarks do GLM-5.2 and o1 share?
28 benchmarks have published results for both models. GLM-5.2 has 51 scored results on Noometry and o1 has 52.