DeepSeek V4 Pro vs GLM-5.2
DeepSeek V4 Pro 
GLM-5.2 Verdict
- DeepSeek V4 Pro (max) and GLM-5.2 (max) are level on quality (64.2 vs 63.9).
- DeepSeek V4 Pro (max) is stronger in agents & tools, composite, instruction following, knowledge, long context.
- GLM-5.2 (max) is stronger in coding, maths, reasoning, human preference.
- DeepSeek V4 Pro (max) is 4.0× cheaper ($0.544 vs $2.15 per 1M blended).
| Metric | DeepSeek V4 Pro (max) | GLM-5.2 (max) |
|---|---|---|
| BenchLeader Index | 64.2 | 63.9 |
| Agents & tools score | 70.2 | 63.8 |
| Coding score | 58.8 | 63.8 |
| Composite score | 75.0 | 72.1 |
| Instruction following score | 74.3 | 71.6 |
| Knowledge score | 58.1 | 55.5 |
| Long context score | 66.6 | 65.6 |
| Maths score | 53.7 | 59.0 |
| Reasoning score | 70.9 | 72.9 |
| Human preference score | – | 67.2 |
| Blended price $/M | $0.544 | $2.15 |
| Output speed | 81 tok/s | 71 tok/s |
| Time to first answer | 55.6 s | 31.6 s |
| Context window | 1M | 1M |
| GPQA Diamond | 89.7% | 91.9% |
| FrontierMath Tiers 1–3 | 45.3% | 59.2% |
| FrontierMath Tier 4 | 2.4% | 29.3% |
| OTIS Mock AIME | 96.7% | 86.4% |
| SWE-bench Verified (Epoch) | 77.6% | 78.7% |
| SimpleQA Verified | 47.0% | 34.2% |
| SciCode | 50.0% | 50.5% |
| WeirdML | 48.9% | 70.1% |
| ProofBench | 16.0% | 35.0% |
| LMArena Text | – | 1472 |
| LMArena Hard Prompts | – | 1493 |
| LMArena Coding | – | 1510 |
| LMArena WebDev | – | 1592 |
| LMArena Agent | – | 4.4 |
| AA Intelligence Index | 36.3 | 34.0 |
| IFBench | 76.5% | 73.3% |
| AA-LCR | 80.3% | 78.3% |
| AA-Omniscience | 0.8 | 4.4 |
| Terminal-Bench Hard | 46.2% | 50.8% |
| GPQA Diamond (AA) | 92.8% | 89.5% |
| Humanity's Last Exam (AA) | 41.0% | 41.1% |
| SciCode (AA) | 51.0% | 51.2% |
| τ²-Bench Telecom (AA) | 96.2% | 99.1% |
| LiveCodeBench | 87.5% | – |
| MMLU-Pro | 87.3% | – |
| LegalBench | 80.3% | – |
| CorpFin | 61.4% | – |
| TaxEval | 72.1% | – |
| Terminal-Bench 2.1 (Vals) | – | 67.8% |
| SWE-bench (Vals) | 77.4% | 82.8% |
| GPQA Diamond (Vals) | 89.4% | – |
| Vals Index | 42.9 | – |
| AIME 2026 | 96.7% | – |
| HMMT February 2026 | 93.9% | – |
| MathArena Apex | 28.1% | – |
| CritPt | 18.0% | 20.9% |
| GDPval (AA) | 49.7% | 45.3% |
| τ²-Bench Banking (AA) | 39.6% | 34.6% |
| ITBench SRE (AA) | 38.3% | 42.7% |
| Analyst Agent (AA) | 18.8% | – |
| APEX-Agents (AA) | 24.3% | 33.7% |
| CaseLaw v2 | 59.4% | – |
| Code Migration | 26.2% | 37.9% |
| Excel Modeling Benchmark | 51.6% | – |
| Finance Agent v2 | 44.1% | – |
| Harvey's Legal Agent Benchmark | 3.8% | 7.1% |
| Legal Research Bench | 23.1% | 31.3% |
| MedCode | 40.5% | – |
| MedScribe | 75.1% | – |
| ProgramBench | 0.0% | 0.5% |
| Public Benefits Bench | 62.9% | – |
| SkillsBench | 51.3% | 45.1% |
| SREBench | – | 0.0% |
| Vibe Code Bench v1.1 | 49.9% | 64.0% |
| LMArena Maths | – | 1480 |
| LMArena Creative Writing | – | 1451 |
| LMArena Instruction Following | – | 1466 |
| LMArena Multi-turn | – | 1469 |
| LMArena Longer Queries | – | 1483 |
| Chess Puzzles | 20.0% | 21.0% |
| EBR-bench | – | 9.5% |
| PostTrainBench | – | 31.7% |
| DeepSWE | – | 43.8% |
| LMCA | 41.2% | 45.8% |
| DTBench | 90.7% | 93.6% |
| CursorBench | – | 55.0% |
| ALE-Bench | – | 1010.2 |
Data as of 2026-09-19. Best configuration of each model; every score links to its source on the model pages.
DeepSeek V4 Pro vs GLM-5.2: questions
- Is DeepSeek V4 Pro better than GLM-5.2?
- DeepSeek V4 Pro (max) and GLM-5.2 (max) are level on quality (64.2 vs 63.9). The BenchLeader Index combines every independent quality benchmark; DeepSeek V4 Pro (max) is ahead overall as of 2026-09-19, but check the category scores for your use.
- Is DeepSeek V4 Pro better than GLM-5.2 for coding?
- GLM-5.2 scores higher in coding (64 vs 59 on the category index, where 50 is average).
- Is DeepSeek V4 Pro better than GLM-5.2 for agentic tasks?
- DeepSeek V4 Pro scores higher in agentic tasks (70 vs 64 on the category index, where 50 is average).
- Which is cheaper, DeepSeek V4 Pro or GLM-5.2?
- DeepSeek V4 Pro is cheaper: $0.544 against $2.15 per million tokens, blended at three input tokens per output token.
- Which is faster, DeepSeek V4 Pro or GLM-5.2?
- DeepSeek V4 Pro streams faster: 81 against 71 output tokens per second.
- Which has the larger context window?
- Both accept 1M tokens of context.