Model comparison
DeepSeek-V2.5 (Sep 2024) vs GLM-4.5-Air
GLM-4.5-Air is the stronger model overall, scoring 38.9 to 37.6 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeekSummary
- They share 17 benchmarks with published results for both. DeepSeek-V2.5 (Sep 2024) scores higher in 1 category and GLM-4.5-Air in 7 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in multilingual, where GLM-4.5-Air leads 49.1 to 42.5.
Side by side
| DeepSeek-V2.5 (Sep 2024) | GLM-4.5-Air | |
|---|---|---|
| Provider | DeepSeek | Z.ai (Zhipu) |
| Noometry Index | 37.6 | 38.9 |
| Released | 2024-09-06 | 2025-07-20 |
| Weights | Open | Open |
| Context window | — | 131K |
| Max output | — | 98K |
| Input $ / M tokens | — | $0.20 |
| Output $ / M tokens | — | $1.10 |
| Results tracked | 22 | 27 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-4.5-Air leads
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), GLM-4.5-Air: 33.3 (#259)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-4.5-Air |
|---|---|---|
| LMArena Coding | 1309 | 1397 |
| Aider Polyglot | 17.8% | — |
| GSO | — | 2.9% |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Reasoning DeepSeek-V2.5 (Sep 2024) leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), GLM-4.5-Air: 24.1 (#166)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-4.5-Air |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1379 |
| Kagi LLM Benchmark | — | 43% |
| ForecastBench | — | 59.2 |
Math Too close to call
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), GLM-4.5-Air: 36.2 (#170)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-4.5-Air |
|---|---|---|
| LMArena Math | 1288 | 1396 |
| Omni-MATH | — | 39.1% |
Knowledge Too close to call
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), GLM-4.5-Air: 35.0 (#191)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-4.5-Air |
|---|---|---|
| LMArena Expert | 1266 | 1370 |
| Humanity's Last Exam | — | 8.1% |
| MMLU-Pro | — | 76.2% |
| Vectara Hallucination Rate | — | 9.3% |
| GPQA (HELM) | — | 59.4% |
Multilingual GLM-4.5-Air leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), GLM-4.5-Air: 49.1 (#135)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-4.5-Air |
|---|---|---|
| LMArena Non-English | 1273 | 1366 |
| LMArena Chinese | 1318 | 1426 |
| LMArena French | 1289 | 1399 |
| LMArena German | 1258 | 1377 |
| LMArena Japanese | 1228 | 1348 |
| LMArena Korean | 1209 | 1308 |
| LMArena Russian | 1289 | 1373 |
| LMArena Spanish | 1248 | 1386 |
Instruction Following GLM-4.5-Air leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), GLM-4.5-Air: 69.6 (#171)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-4.5-Air |
|---|---|---|
| LMArena Instruction Following | 1280 | 1354 |
| IFEval | — | 81.2% |
Long Context GLM-4.5-Air leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), GLM-4.5-Air: 41.6 (#135)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-4.5-Air |
|---|---|---|
| LMArena Longer Query | 1301 | 1366 |
Writing & Preference GLM-4.5-Air leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), GLM-4.5-Air: 55.9 (#139)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-4.5-Air |
|---|---|---|
| LMArena Text | 1294 | 1384 |
| LMArena Creative Writing | 1285 | 1343 |
| LMArena Multi-Turn | 1297 | 1371 |
| WildBench | — | 78.9% |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than GLM-4.5-Air?
GLM-4.5-Air is the stronger model overall, scoring 38.9 to 37.6 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or GLM-4.5-Air better for coding?
GLM-4.5-Air scores higher on coding benchmarks: 33.3 versus 31.7 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and GLM-4.5-Air share?
17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and GLM-4.5-Air has 27.