Model comparison
Deepseek Coder v2 vs GLM-4.6
GLM-4.6 is the stronger model overall, scoring 40.9 to 35.3 on the Noometry Index.
Last verified . 17 shared benchmarks.
DeepSeekSummary
- They share 17 benchmarks with published results for both. Deepseek Coder v2 scores higher in 0 categories and GLM-4.6 in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where GLM-4.6 leads 61.3 to 38.4.
Side by side
| Deepseek Coder v2 | GLM-4.6 | |
|---|---|---|
| Provider | DeepSeek | Z.ai (Zhipu) |
| Noometry Index | 35.3 | 40.9 |
| Released | 2024-06-17 | 2025-09-30 |
| Weights | Open | Open |
| Context window | — | 205K |
| Max output | — | 131K |
| Input $ / M tokens | — | $0.60 |
| Output $ / M tokens | — | $2.20 |
| Results tracked | 24 | 29 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-4.6 leads
Deepseek Coder v2: 37.5 (#178), GLM-4.6: 39.9 (#137)
| Benchmark | Deepseek Coder v2 | GLM-4.6 |
|---|---|---|
| LMArena Coding | 1251 | 1449 |
| SWE-bench Verified (bash only) | — | 55.4% |
| LMArena WebDev | — | 1339 |
| SciCode | — | 38.4% |
| BigCodeBench Instruct | 48.2% | — |
| BigCodeBench Complete | 59.7% | — |
| ALE-Bench | — | 340.82 |
| HumanEval+ | 82.3% | — |
| MBPP+ | 75.1% | — |
Agentic & Tool Use Not comparable
Deepseek Coder v2: —, GLM-4.6: 32.1 (#60)
| Benchmark | Deepseek Coder v2 | GLM-4.6 |
|---|---|---|
| Terminal-Bench | — | 24.5% |
| Berkeley Function Calling Leaderboard | — | 72.4% |
Reasoning Too close to call
Deepseek Coder v2: 22.6 (#175), GLM-4.6: 23.1 (#168)
| Benchmark | Deepseek Coder v2 | GLM-4.6 |
|---|---|---|
| LMArena Hard Prompts | 1207 | 1440 |
| Kagi LLM Benchmark | — | 47.4% |
| CritPt | — | 1.1% |
| WinoGrande | 83.7% | — |
Math GLM-4.6 leads
Deepseek Coder v2: 32.3 (#188), GLM-4.6: 36.4 (#115)
| Benchmark | Deepseek Coder v2 | GLM-4.6 |
|---|---|---|
| LMArena Math | 1241 | 1431 |
| FrontierMath (Feb 2025 set) | — | 3.8% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
| GSM8K | 94.5% | — |
Knowledge GLM-4.6 leads
Deepseek Coder v2: 31.7 (#209), GLM-4.6: 39.6 (#118)
| Benchmark | Deepseek Coder v2 | GLM-4.6 |
|---|---|---|
| LMArena Expert | 1181 | 1432 |
| Vectara Hallucination Rate | — | 9.5% |
| ARC (AI2) Challenge | 64.3% | — |
Multilingual GLM-4.6 leads
Deepseek Coder v2: 36.4 (#235), GLM-4.6: 53.6 (#62)
| Benchmark | Deepseek Coder v2 | GLM-4.6 |
|---|---|---|
| LMArena Non-English | 1182 | 1426 |
| LMArena Chinese | 1201 | 1500 |
| LMArena French | 1184 | 1459 |
| LMArena German | 1164 | 1447 |
| LMArena Japanese | 1126 | 1394 |
| LMArena Korean | 1104 | 1401 |
| LMArena Russian | 1188 | 1418 |
| LMArena Spanish | 1153 | 1436 |
Instruction Following GLM-4.6 leads
Deepseek Coder v2: 61.9 (#237), GLM-4.6: 74.5 (#93)
| Benchmark | Deepseek Coder v2 | GLM-4.6 |
|---|---|---|
| LMArena Instruction Following | 1180 | 1410 |
Long Context GLM-4.6 leads
Deepseek Coder v2: 37.0 (#219), GLM-4.6: 43.4 (#91)
| Benchmark | Deepseek Coder v2 | GLM-4.6 |
|---|---|---|
| LMArena Longer Query | 1219 | 1423 |
Writing & Preference GLM-4.6 leads
Deepseek Coder v2: 38.4 (#249), GLM-4.6: 61.3 (#86)
| Benchmark | Deepseek Coder v2 | GLM-4.6 |
|---|---|---|
| LMArena Text | 1191 | 1440 |
| LMArena Creative Writing | 1120 | 1411 |
| LMArena Multi-Turn | 1176 | 1427 |
| EQ-Bench Creative Writing | — | 1411 |
Frequently asked questions
Is Deepseek Coder v2 better than GLM-4.6?
GLM-4.6 is the stronger model overall, scoring 40.9 to 35.3 on the Noometry Index.
Is Deepseek Coder v2 or GLM-4.6 better for coding?
GLM-4.6 scores higher on coding benchmarks: 39.9 versus 37.5 in the Noometry coding category.
How many benchmarks do Deepseek Coder v2 and GLM-4.6 share?
17 benchmarks have published results for both models. Deepseek Coder v2 has 24 scored results on Noometry and GLM-4.6 has 29.