Model comparison
GLM-4.5-Air vs Grok-2 (Dec 2024)
GLM-4.5-Air is the stronger model overall, scoring 38.9 to 33.7 on the Noometry Index.
Last verified . 17 shared benchmarks.
Summary
- They share 17 benchmarks with published results for both. GLM-4.5-Air scores higher in 7 categories and Grok-2 (Dec 2024) in 1 category; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-4.5-Air leads 36.2 to 20.8.
- GLM-4.5-Air has downloadable open weights; the other is API-only.
Side by side
| GLM-4.5-Air | Grok-2 (Dec 2024) | |
|---|---|---|
| Provider | Z.ai (Zhipu) | xAI |
| Noometry Index | 38.9 | 33.7 |
| Released | 2025-07-20 | 2024-08-13 |
| Weights | Open | Proprietary |
| Context window | 131K | — |
| Max output | 98K | — |
| Input $ / M tokens | $0.20 | — |
| Output $ / M tokens | $1.10 | — |
| Results tracked | 27 | 34 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
GLM-4.5-Air: 33.3 (#259), Grok-2 (Dec 2024): 33.3 (#258)
| Benchmark | GLM-4.5-Air | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Coding | 1397 | 1287 |
| GSO | 2.9% | — |
| WeirdML | — | 22.2% |
| LiveBench Coding | — | 46.4% |
Reasoning GLM-4.5-Air leads
GLM-4.5-Air: 24.1 (#166), Grok-2 (Dec 2024): 16.9 (#299)
| Benchmark | GLM-4.5-Air | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Hard Prompts | 1379 | 1272 |
| SimpleBench | — | 22.7% |
| Kagi LLM Benchmark | 43% | — |
| LiveBench Reasoning | — | 54.8% |
| DTBench | — | 65.2% |
| LiveBench Data Analysis | — | 54.5% |
| Epoch Capabilities Index | — | 130.48 |
| ForecastBench | 59.2 | — |
| LiveBench | — | 54.3% |
Math GLM-4.5-Air leads
GLM-4.5-Air: 36.2 (#170), Grok-2 (Dec 2024): 20.8 (#284)
| Benchmark | GLM-4.5-Air | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Math | 1396 | 1283 |
| OTIS Mock AIME 2024-2025 | — | 11.5% |
| Omni-MATH | 39.1% | — |
| LiveBench Math | — | 54.9% |
| MATH Level 5 | — | 63.5% |
| FrontierMath (Feb 2025 set) | — | 0.7% |
Knowledge GLM-4.5-Air leads
GLM-4.5-Air: 35.0 (#191), Grok-2 (Dec 2024): 29.8 (#233)
| Benchmark | GLM-4.5-Air | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Expert | 1370 | 1254 |
| GPQA Diamond | — | 53.8% |
| Humanity's Last Exam | 8.1% | — |
| MMLU-Pro | 76.2% | — |
| Confabulations | — | 20.1% |
| Vectara Hallucination Rate | 9.3% | — |
| GPQA (HELM) | 59.4% | — |
Multilingual GLM-4.5-Air leads
GLM-4.5-Air: 49.1 (#135), Grok-2 (Dec 2024): 43.1 (#188)
| Benchmark | GLM-4.5-Air | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Non-English | 1366 | 1282 |
| LMArena Chinese | 1426 | 1289 |
| LMArena French | 1399 | 1318 |
| LMArena German | 1377 | 1287 |
| LMArena Japanese | 1348 | 1244 |
| LMArena Korean | 1308 | 1237 |
| LMArena Russian | 1373 | 1286 |
| LMArena Spanish | 1386 | 1281 |
Instruction Following GLM-4.5-Air leads
GLM-4.5-Air: 69.6 (#171), Grok-2 (Dec 2024): 66.9 (#202)
| Benchmark | GLM-4.5-Air | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Instruction Following | 1354 | 1270 |
| LiveBench Instruction Following | — | 69.6% |
| IFEval | 81.2% | — |
Long Context GLM-4.5-Air leads
GLM-4.5-Air: 41.6 (#135), Grok-2 (Dec 2024): 38.8 (#190)
| Benchmark | GLM-4.5-Air | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Longer Query | 1366 | 1276 |
Writing & Preference GLM-4.5-Air leads
GLM-4.5-Air: 55.9 (#139), Grok-2 (Dec 2024): 48.6 (#198)
| Benchmark | GLM-4.5-Air | Grok-2 (Dec 2024) |
|---|---|---|
| LMArena Text | 1384 | 1305 |
| LMArena Creative Writing | 1343 | 1284 |
| LMArena Multi-Turn | 1371 | 1290 |
| Short-Story Creative Writing | — | 63.6% |
| WildBench | 78.9% | — |
| LiveBench Language | — | 45.6% |
Frequently asked questions
Is GLM-4.5-Air better than Grok-2 (Dec 2024)?
GLM-4.5-Air is the stronger model overall, scoring 38.9 to 33.7 on the Noometry Index.
Is GLM-4.5-Air or Grok-2 (Dec 2024) better for coding?
They score almost the same on coding (33.3 vs 33.3); test both on your own repository before choosing.
How many benchmarks do GLM-4.5-Air and Grok-2 (Dec 2024) share?
17 benchmarks have published results for both models. GLM-4.5-Air has 27 scored results on Noometry and Grok-2 (Dec 2024) has 34.