Model comparison
DeepSeek-V2.5 (Sep 2024) vs GLM-5
GLM-5 is the stronger model overall, scoring 46.1 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 0 categories and GLM-5 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where GLM-5 leads 52.3 to 34.8.
Side by side
| DeepSeek-V2.5 (Sep 2024) | GLM-5 | |
|---|---|---|
| Provider | DeepSeek | Z.ai (Zhipu) |
| Noometry Index | 37.6 | 46.1 |
| Released | 2024-09-06 | 2026-02-11 |
| Weights | Open | Open |
| Context window | — | 205K |
| Max output | — | 131K |
| Input $ / M tokens | — | $1 |
| Output $ / M tokens | — | $3.20 |
| Results tracked | 22 | 45 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5 leads
DeepSeek-V2.5 (Sep 2024): 31.7 (#281), GLM-5: 49.0 (#52)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-5 |
|---|---|---|
| LMArena Coding | 1309 | 1461 |
| SWE-bench Verified | — | 72.1% |
| SWE-bench Verified (bash only) | — | 72.8% |
| Aider Polyglot | 17.8% | — |
| LMArena WebDev | — | 1434 |
| SWE-bench Multilingual | — | 69.7% |
| WeirdML | — | 48.2% |
| BigCodeBench Instruct | 48.6% | — |
| BigCodeBench Complete | 53.2% | — |
| ALE-Bench | — | 765.62 |
| HumanEval+ | 83.5% | — |
| MBPP+ | 74.1% | — |
Agentic & Tool Use Not comparable
DeepSeek-V2.5 (Sep 2024): —, GLM-5: 31.1 (#71)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-5 |
|---|---|---|
| Terminal-Bench | — | 52.4% |
| τ²-bench Airline | — | 82.5% |
| τ²-bench Banking | — | 9.8% |
| τ²-bench Retail | — | 73.7% |
| τ²-bench Telecom | — | 86.8% |
| Vending-Bench 2 | — | 4,432 |
Reasoning GLM-5 leads
DeepSeek-V2.5 (Sep 2024): 25.6 (#145), GLM-5: 27.6 (#116)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-5 |
|---|---|---|
| LMArena Hard Prompts | 1289 | 1452 |
| ARC-AGI-2 | — | 4.9% |
| SimpleBench | — | 53.2% |
| Kagi LLM Benchmark | — | 75% |
| NYT Connections (extended) | — | 74.8% |
| ARC-AGI-1 | — | 44.7% |
| Chess Puzzles | — | 10% |
| Epoch Capabilities Index | — | 145.83 |
| ForecastBench | — | 61 |
Math GLM-5 leads
DeepSeek-V2.5 (Sep 2024): 35.9 (#177), GLM-5: 46.4 (#71)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-5 |
|---|---|---|
| LMArena Math | 1288 | 1440 |
| MathArena Final-Answer Competitions | — | 65.7% |
| OTIS Mock AIME 2024-2025 | — | 80% |
| FrontierMath (Feb 2025 set) | — | 16.4% |
| FrontierMath Tier 4 (v1) | — | 2.1% |
Knowledge GLM-5 leads
DeepSeek-V2.5 (Sep 2024): 34.8 (#193), GLM-5: 52.3 (#64)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-5 |
|---|---|---|
| LMArena Expert | 1266 | 1454 |
| GPQA Diamond | — | 87.8% |
| Vectara Hallucination Rate | — | 10.1% |
Multilingual GLM-5 leads
DeepSeek-V2.5 (Sep 2024): 42.5 (#193), GLM-5: 53.7 (#58)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-5 |
|---|---|---|
| LMArena Non-English | 1273 | 1430 |
| LMArena Chinese | 1318 | 1511 |
| LMArena French | 1289 | 1455 |
| LMArena German | 1258 | 1445 |
| LMArena Japanese | 1228 | 1416 |
| LMArena Korean | 1209 | 1423 |
| LMArena Russian | 1289 | 1436 |
| LMArena Spanish | 1248 | 1454 |
Instruction Following GLM-5 leads
DeepSeek-V2.5 (Sep 2024): 67.5 (#194), GLM-5: 75.2 (#67)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-5 |
|---|---|---|
| LMArena Instruction Following | 1280 | 1428 |
Long Context GLM-5 leads
DeepSeek-V2.5 (Sep 2024): 39.5 (#174), GLM-5: 44.7 (#60)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-5 |
|---|---|---|
| LMArena Longer Query | 1301 | 1446 |
| CL-bench | — | 18.7% |
Writing & Preference GLM-5 leads
DeepSeek-V2.5 (Sep 2024): 49.8 (#187), GLM-5: 66.0 (#38)
| Benchmark | DeepSeek-V2.5 (Sep 2024) | GLM-5 |
|---|---|---|
| LMArena Text | 1294 | 1446 |
| LMArena Creative Writing | 1285 | 1439 |
| LMArena Multi-Turn | 1297 | 1456 |
| EQ-Bench Creative Writing | — | 1601 |
Frequently asked questions
Is DeepSeek-V2.5 (Sep 2024) better than GLM-5?
GLM-5 is the stronger model overall, scoring 46.1 to 37.6 on the Noometry Index.
Is DeepSeek-V2.5 (Sep 2024) or GLM-5 better for coding?
GLM-5 scores higher on coding benchmarks: 49.0 versus 31.7 in the Noometry coding category.
How many benchmarks do DeepSeek-V2.5 (Sep 2024) and GLM-5 share?
17 benchmarks have published results for both models. DeepSeek-V2.5 (Sep 2024) has 22 scored results on Noometry and GLM-5 has 45.