Model comparison
DeepSeek-V3 vs GLM-5.1
GLM-5.1 is the stronger model overall, scoring 47.6 to 39.2 on the Noometry Index. DeepSeek-V3 costs 5.3× less per token, which makes it the better buy when GLM-5.1's lead doesn't matter for your workload.
Last verified . 26 shared benchmarks.
DeepSeekSummary
- They share 26 benchmarks with published results for both. DeepSeek-V3 scores higher in 0 categories and GLM-5.1 in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in math, where GLM-5.1 leads 49.0 to 30.4.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 37.8% for DeepSeek-V3 and 93.3% for GLM-5.1.
- DeepSeek-V3 is cheaper at $0.24 / $0.90 per million input/output tokens, against $1.40 / $4.40 for GLM-5.1.
- GLM-5.1 accepts more context: 200K tokens versus 164K.
Side by side
| DeepSeek-V3 | GLM-5.1 | |
|---|---|---|
| Provider | DeepSeek | Z.ai (Zhipu) |
| Noometry Index | 39.2 | 47.6 |
| Released | 2024-12-26 | 2026-04-07 |
| Weights | Open | Open |
| Context window | 164K | 200K |
| Max output | 164K | 131K |
| Input $ / M tokens | $0.24 | $1.40 |
| Output $ / M tokens | $0.90 | $4.40 |
| Results tracked | 60 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding GLM-5.1 leads
DeepSeek-V3: 41.9 (#99), GLM-5.1: 48.5 (#52)
| Benchmark | DeepSeek-V3 | GLM-5.1 |
|---|---|---|
| SciCode | 35.8% | 43.8% |
| WeirdML | 36.1% | 57.1% |
| LMArena Coding | 1368 | 1486 |
| SWE-bench Verified | — | 74.2% |
| Aider Polyglot | 55.1% | — |
| LMArena WebDev | — | 1509 |
| BigCodeBench Instruct | 50% | — |
| LiveBench Coding | 70.9% | — |
| BigCodeBench Complete | 62.2% | — |
| ALE-Bench | — | 887.1 |
| HumanEval+ | 86.6% | — |
| MBPP+ | 73% | — |
Agentic & Tool Use Not comparable
DeepSeek-V3: —, GLM-5.1: 24.8 (#106)
| Benchmark | DeepSeek-V3 | GLM-5.1 |
|---|---|---|
| APEX-Agents | — | 40.9% |
| ExploitBench | — | 18.1% |
| GBAEval | — | 0% |
| METR Time Horizons | 49.6% | — |
| Vending-Bench 2 | — | 5,634 |
Reasoning GLM-5.1 leads
DeepSeek-V3: 20.0 (#226), GLM-5.1: 38.2 (#57)
| Benchmark | DeepSeek-V3 | GLM-5.1 |
|---|---|---|
| SimpleBench | 27.2% | 55.1% |
| CritPt | 0% | 4.6% |
| LMArena Hard Prompts | 1365 | 1472 |
| Epoch Capabilities Index | 135.95 | 149.86 |
| Kagi LLM Benchmark | 52.3% | — |
| NYT Connections (extended) | — | 77.7% |
| Chess Puzzles | — | 19% |
| Thematic Generalization | — | 69.8% |
| LiveBench Reasoning | 65.8% | — |
| DTBench | 64.8% | — |
| LiveBench Data Analysis | 60.9% | — |
| LMCA | 15.5% | — |
| BIG-Bench Hard | 87.5% | — |
| ForecastBench | 59.1 | — |
| HellaSwag | 88.9% | — |
| LiveBench | 66.9% | — |
| PIQA | 84.7% | — |
| WinoGrande | 85.2% | — |
Math GLM-5.1 leads
DeepSeek-V3: 30.4 (#208), GLM-5.1: 49.0 (#48)
| Benchmark | DeepSeek-V3 | GLM-5.1 |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 37.8% | 93.3% |
| LMArena Math | 1373 | 1472 |
| FrontierMath (Feb 2025 set) | 1.7% | 33.4% |
| FrontierMath (Tiers 1-3) | — | 36.8% |
| MathArena Final-Answer Competitions | — | 67.1% |
| ProofBench | — | 22.2% |
| Omni-MATH | 40.3% | — |
| LiveBench Math | 73.5% | — |
| MATH Level 5 | 75.5% | — |
| FrontierMath Tier 4 (v1) | — | 12.5% |
Knowledge GLM-5.1 leads
DeepSeek-V3: 37.5 (#147), GLM-5.1: 55.0 (#45)
| Benchmark | DeepSeek-V3 | GLM-5.1 |
|---|---|---|
| GPQA Diamond | 67.6% | 89.9% |
| LMArena Expert | 1352 | 1477 |
| SimpleQA Verified | — | 34% |
| MMLU-Pro | 72.3% | — |
| Confabulations | 26.1% | — |
| Vectara Hallucination Rate | 6.1% | — |
| GPQA (HELM) | 53.8% | — |
| ARC (AI2) Challenge | 95.3% | — |
| MMLU | 87.2% | — |
| TriviaQA | 82.9% | — |
Multilingual GLM-5.1 leads
DeepSeek-V3: 48.8 (#138), GLM-5.1: 55.2 (#34)
| Benchmark | DeepSeek-V3 | GLM-5.1 |
|---|---|---|
| LMArena Non-English | 1358 | 1448 |
| LMArena Chinese | 1390 | 1516 |
| LMArena French | 1386 | 1478 |
| LMArena German | 1374 | 1466 |
| LMArena Japanese | 1333 | 1436 |
| LMArena Korean | 1321 | 1423 |
| LMArena Russian | 1372 | 1453 |
| LMArena Spanish | 1358 | 1467 |
Instruction Following GLM-5.1 leads
DeepSeek-V3: 73.0 (#125), GLM-5.1: 76.5 (#41)
| Benchmark | DeepSeek-V3 | GLM-5.1 |
|---|---|---|
| LMArena Instruction Following | 1345 | 1451 |
| LiveBench Instruction Following | 81.5% | — |
| IFEval | 83.2% | — |
Long Context GLM-5.1 leads
DeepSeek-V3: 34.0 (#248), GLM-5.1: 44.9 (#51)
| Benchmark | DeepSeek-V3 | GLM-5.1 |
|---|---|---|
| LMArena Longer Query | 1353 | 1467 |
| Fiction.LiveBench | 50% | — |
Writing & Preference GLM-5.1 leads
DeepSeek-V3: 57.6 (#124), GLM-5.1: 67.2 (#29)
| Benchmark | DeepSeek-V3 | GLM-5.1 |
|---|---|---|
| LMArena Text | 1375 | 1462 |
| LMArena Creative Writing | 1364 | 1454 |
| EQ-Bench Creative Writing | 1472 | 1592 |
| LMArena Multi-Turn | 1389 | 1473 |
| Short-Story Creative Writing | 77% | — |
| WildBench | 83% | — |
| LiveBench Language | 49.1% | — |
Frequently asked questions
Is DeepSeek-V3 better than GLM-5.1?
GLM-5.1 is the stronger model overall, scoring 47.6 to 39.2 on the Noometry Index. DeepSeek-V3 costs 5.3× less per token, which makes it the better buy when GLM-5.1's lead doesn't matter for your workload.
Which is cheaper, DeepSeek-V3 or GLM-5.1?
DeepSeek-V3 is cheaper. It lists at $0.24 per million input tokens and $0.90 per million output tokens; GLM-5.1 lists at $1.40 and $4.40.
Is DeepSeek-V3 or GLM-5.1 better for coding?
GLM-5.1 scores higher on coding benchmarks: 48.5 versus 41.9 in the Noometry coding category.
Which has the bigger context window?
GLM-5.1 does, with 200K tokens against 164K.
How many benchmarks do DeepSeek-V3 and GLM-5.1 share?
26 benchmarks have published results for both models. DeepSeek-V3 has 60 scored results on Noometry and GLM-5.1 has 41.