Model comparison
GLM-5.2 vs Grok 4.5
Grok 4.5 is the stronger model overall, scoring 55.0 to 51.1 on the Noometry Index.
Last verified . 46 shared benchmarks.
Summary
- They share 46 benchmarks with published results for both. GLM-5.2 scores higher in 4 categories and Grok 4.5 in 5 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.5 leads 56.1 to 42.3.
- The biggest single-benchmark swing is GBAEval: 0% for GLM-5.2 and 65.4% for Grok 4.5.
- GLM-5.2 is cheaper at $1.40 / $4.40 per million input/output tokens, against $2 / $6 for Grok 4.5.
- GLM-5.2 accepts more context: 1M tokens versus 500K.
- GLM-5.2 has downloadable open weights; the other is API-only.
Side by side
| GLM-5.2 | Grok 4.5 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | xAI |
| Noometry Index | 51.1 | 55.0 |
| Released | 2026-06-13 | 2026-07-08 |
| Weights | Open | Proprietary |
| Context window | 1M | 500K |
| Max output | 131K | 500K |
| Input $ / M tokens | $1.40 | $2 |
| Output $ / M tokens | $4.40 | $6 |
| Results tracked | 51 | 52 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GLM-5.2: 51.3 (#41), Grok 4.5: 52.2 (#35)
| Benchmark | GLM-5.2 | Grok 4.5 |
|---|---|---|
| DeepSWE | 43.8% | 53.8% |
| FrontierCode | 24.5% | 42.4% |
| LMArena WebDev | 1603 | 1553 |
| SciCode | 50.5% | 54.1% |
| WeirdML | 70.1% | 46.4% |
| LMArena Coding | 1485 | 1474 |
| ALE-Bench | 1,047 | 1,309 |
| SWE-bench Verified | 78.7% | — |
Agentic & Tool Use Grok 4.5 leads
GLM-5.2: 32.4 (#63), Grok 4.5: 44.4 (#17)
| Benchmark | GLM-5.2 | Grok 4.5 |
|---|---|---|
| APEX-Agents | 45.2% | 56.2% |
| τ²-bench Banking | 37.1% | 47.9% |
| PostTrainBench | 31.7% | 23.4% |
| GBAEval | 0% | 65.4% |
| Vending-Bench 2 | 8,314 | 3,887 |
| GDP.pdf | — | 14% |
| LMArena Search | — | 1213 |
Reasoning Grok 4.5 leads
GLM-5.2: 42.3 (#52), Grok 4.5: 56.1 (#25)
| Benchmark | GLM-5.2 | Grok 4.5 |
|---|---|---|
| ARC-AGI-2 | 22.8% | 52.6% |
| SimpleBench | 58.8% | 70% |
| Kagi LLM Benchmark | 62.6% | 83.5% |
| NYT Connections (extended) | 74.3% | 79.9% |
| ARC-AGI-1 | 77% | 87.2% |
| CritPt | 20.9% | 15.4% |
| Chess Puzzles | 21% | 36% |
| LMArena Hard Prompts | 1480 | 1462 |
| DTBench | 93.6% | 96.5% |
| LMCA | 45.8% | 45.2% |
| Surface Evolver Bench | 55.6% | 74.4% |
| Epoch Capabilities Index | 151.78 | 153.92 |
| EBR-Bench | 9.5% | — |
| Mystery Game Puzzles | 19% | — |
Math Grok 4.5 leads
GLM-5.2: 55.7 (#43), Grok 4.5: 60.9 (#35)
| Benchmark | GLM-5.2 | Grok 4.5 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 59.2% | 57.2% |
| FrontierMath Tier 4 | 29.3% | 24.4% |
| OTIS Mock AIME 2024-2025 | 86.4% | 97.8% |
| ProofBench | 35% | 31% |
| LMArena Math | 1482 | 1459 |
| MathArena Final-Answer Competitions | 67.6% | — |
Knowledge Grok 4.5 leads
GLM-5.2: 57.1 (#40), Grok 4.5: 62.3 (#24)
| Benchmark | GLM-5.2 | Grok 4.5 |
|---|---|---|
| GPQA Diamond | 91.9% | 93.4% |
| SimpleQA Verified | 34.2% | 48.3% |
| LMArena Expert | 1486 | 1466 |
Multimodal Not comparable
GLM-5.2: —, Grok 4.5: 37.6 (#72)
| Benchmark | GLM-5.2 | Grok 4.5 |
|---|---|---|
| LMArena Vision | — | 1288 |
| Blueprint-Bench 2 | — | 27.3% |
| Furniture Assembly | — | 22.5% |
| LMArena Document | — | 1452 |
Multilingual GLM-5.2 leads
GLM-5.2: 55.8 (#26), Grok 4.5: 54.4 (#42)
| Benchmark | GLM-5.2 | Grok 4.5 |
|---|---|---|
| LMArena Non-English | 1459 | 1440 |
| LMArena Chinese | 1519 | 1496 |
| LMArena French | 1479 | 1456 |
| LMArena German | 1468 | 1446 |
| LMArena Japanese | 1451 | 1428 |
| LMArena Korean | 1445 | 1404 |
| LMArena Russian | 1466 | 1448 |
| LMArena Spanish | 1477 | 1450 |
Instruction Following Too close to call
GLM-5.2: 76.9 (#34), Grok 4.5: 76.0 (#48)
| Benchmark | GLM-5.2 | Grok 4.5 |
|---|---|---|
| LMArena Instruction Following | 1465 | 1446 |
Long Context Too close to call
GLM-5.2: 45.3 (#43), Grok 4.5: 44.8 (#56)
| Benchmark | GLM-5.2 | Grok 4.5 |
|---|---|---|
| LMArena Longer Query | 1479 | 1463 |
Writing & Preference GLM-5.2 leads
GLM-5.2: 70.4 (#21), Grok 4.5: 65.8 (#42)
| Benchmark | GLM-5.2 | Grok 4.5 |
|---|---|---|
| LMArena Text | 1470 | 1448 |
| LMArena Creative Writing | 1462 | 1442 |
| EQ-Bench Creative Writing | 1757 | 1579 |
| LMArena Multi-Turn | 1469 | 1456 |
| EQ-Bench 4 | 1222 | — |
Frequently asked questions
Is GLM-5.2 better than Grok 4.5?
Grok 4.5 is the stronger model overall, scoring 55.0 to 51.1 on the Noometry Index.
Which is cheaper, GLM-5.2 or Grok 4.5?
GLM-5.2 is cheaper. It lists at $1.40 per million input tokens and $4.40 per million output tokens; Grok 4.5 lists at $2 and $6.
Is GLM-5.2 or Grok 4.5 better for coding?
They score almost the same on coding (51.3 vs 52.2); test both on your own repository before choosing.
Which has the bigger context window?
GLM-5.2 does, with 1M tokens against 500K.
How many benchmarks do GLM-5.2 and Grok 4.5 share?
46 benchmarks have published results for both models. GLM-5.2 has 51 scored results on Noometry and Grok 4.5 has 52.