Model comparison
GLM-5 vs GLM-5.2
GLM-5.2 is the stronger model overall, scoring 51.1 to 46.1 on the Noometry Index.
Last verified . 34 shared benchmarks.
Summary
- They share 34 benchmarks with published results for both. GLM-5 scores higher in 0 categories and GLM-5.2 in 9 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where GLM-5.2 leads 42.3 to 27.6.
- The biggest single-benchmark swing is ARC-AGI-1: 44.7% for GLM-5 and 77% for GLM-5.2.
- GLM-5 is cheaper at $1 / $3.20 per million input/output tokens, against $1.40 / $4.40 for GLM-5.2.
- GLM-5.2 accepts more context: 1M tokens versus 205K.
Side by side
| GLM-5 | GLM-5.2 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | Z.ai (Zhipu) |
| Noometry Index | 46.1 | 51.1 |
| Released | 2026-02-11 | 2026-06-13 |
| Weights | Open | Open |
| Context window | 205K | 1M |
| Max output | 131K | 131K |
| Input $ / M tokens | $1 | $1.40 |
| Output $ / M tokens | $3.20 | $4.40 |
| Results tracked | 45 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.2 leads
GLM-5: 49.0 (#52), GLM-5.2: 51.3 (#41)
| Benchmark | GLM-5 | GLM-5.2 |
|---|---|---|
| SWE-bench Verified | 72.1% | 78.7% |
| LMArena WebDev | 1434 | 1603 |
| WeirdML | 48.2% | 70.1% |
| LMArena Coding | 1461 | 1485 |
| ALE-Bench | 765.62 | 1,047 |
| DeepSWE | — | 43.8% |
| FrontierCode | — | 24.5% |
| SWE-bench Verified (bash only) | 72.8% | — |
| SWE-bench Multilingual | 69.7% | — |
| SciCode | — | 50.5% |
Agentic & Tool Use GLM-5.2 leads
GLM-5: 31.1 (#71), GLM-5.2: 32.4 (#63)
| Benchmark | GLM-5 | GLM-5.2 |
|---|---|---|
| τ²-bench Banking | 9.8% | 37.1% |
| Vending-Bench 2 | 4,432 | 8,314 |
| Terminal-Bench | 52.4% | — |
| APEX-Agents | — | 45.2% |
| τ²-bench Airline | 82.5% | — |
| τ²-bench Retail | 73.7% | — |
| τ²-bench Telecom | 86.8% | — |
| PostTrainBench | — | 31.7% |
| GBAEval | — | 0% |
Reasoning GLM-5.2 leads
GLM-5: 27.6 (#116), GLM-5.2: 42.3 (#52)
| Benchmark | GLM-5 | GLM-5.2 |
|---|---|---|
| ARC-AGI-2 | 4.9% | 22.8% |
| SimpleBench | 53.2% | 58.8% |
| Kagi LLM Benchmark | 75% | 62.6% |
| NYT Connections (extended) | 74.8% | 74.3% |
| ARC-AGI-1 | 44.7% | 77% |
| Chess Puzzles | 10% | 21% |
| LMArena Hard Prompts | 1452 | 1480 |
| Epoch Capabilities Index | 145.83 | 151.78 |
| CritPt | — | 20.9% |
| EBR-Bench | — | 9.5% |
| Mystery Game Puzzles | — | 19% |
| DTBench | — | 93.6% |
| LMCA | — | 45.8% |
| Surface Evolver Bench | — | 55.6% |
| ForecastBench | 61 | — |
Math GLM-5.2 leads
GLM-5: 46.4 (#71), GLM-5.2: 55.7 (#43)
| Benchmark | GLM-5 | GLM-5.2 |
|---|---|---|
| MathArena Final-Answer Competitions | 65.7% | 67.6% |
| OTIS Mock AIME 2024-2025 | 80% | 86.4% |
| LMArena Math | 1440 | 1482 |
| FrontierMath (Tiers 1-3) | — | 59.2% |
| FrontierMath Tier 4 | — | 29.3% |
| ProofBench | — | 35% |
| FrontierMath (Feb 2025 set) | 16.4% | — |
| FrontierMath Tier 4 (v1) | 2.1% | — |
Knowledge GLM-5.2 leads
GLM-5: 52.3 (#64), GLM-5.2: 57.1 (#40)
| Benchmark | GLM-5 | GLM-5.2 |
|---|---|---|
| GPQA Diamond | 87.8% | 91.9% |
| LMArena Expert | 1454 | 1486 |
| SimpleQA Verified | — | 34.2% |
| Vectara Hallucination Rate | 10.1% | — |
Multilingual GLM-5.2 leads
GLM-5: 53.7 (#58), GLM-5.2: 55.8 (#26)
| Benchmark | GLM-5 | GLM-5.2 |
|---|---|---|
| LMArena Non-English | 1430 | 1459 |
| LMArena Chinese | 1511 | 1519 |
| LMArena French | 1455 | 1479 |
| LMArena German | 1445 | 1468 |
| LMArena Japanese | 1416 | 1451 |
| LMArena Korean | 1423 | 1445 |
| LMArena Russian | 1436 | 1466 |
| LMArena Spanish | 1454 | 1477 |
Instruction Following GLM-5.2 leads
GLM-5: 75.2 (#67), GLM-5.2: 76.9 (#34)
| Benchmark | GLM-5 | GLM-5.2 |
|---|---|---|
| LMArena Instruction Following | 1428 | 1465 |
Long Context Too close to call
GLM-5: 44.7 (#60), GLM-5.2: 45.3 (#43)
| Benchmark | GLM-5 | GLM-5.2 |
|---|---|---|
| LMArena Longer Query | 1446 | 1479 |
| CL-bench | 18.7% | — |
Writing & Preference GLM-5.2 leads
GLM-5: 66.0 (#38), GLM-5.2: 70.4 (#21)
| Benchmark | GLM-5 | GLM-5.2 |
|---|---|---|
| LMArena Text | 1446 | 1470 |
| LMArena Creative Writing | 1439 | 1462 |
| EQ-Bench Creative Writing | 1601 | 1757 |
| LMArena Multi-Turn | 1456 | 1469 |
| EQ-Bench 4 | — | 1222 |
Frequently asked questions
Is GLM-5 better than GLM-5.2?
GLM-5.2 is the stronger model overall, scoring 51.1 to 46.1 on the Noometry Index.
Which is cheaper, GLM-5 or GLM-5.2?
GLM-5 is cheaper. It lists at $1 per million input tokens and $3.20 per million output tokens; GLM-5.2 lists at $1.40 and $4.40.
Is GLM-5 or GLM-5.2 better for coding?
GLM-5.2 scores higher on coding benchmarks: 51.3 versus 49.0 in the Noometry coding category.
Which has the bigger context window?
GLM-5.2 does, with 1M tokens against 205K.
How many benchmarks do GLM-5 and GLM-5.2 share?
34 benchmarks have published results for both models. GLM-5 has 45 scored results on Noometry and GLM-5.2 has 51.