GLM-5.2 vs GPT-5.4
GLM-5.2 
GPT-5.4 Verdict
- GPT-5.4 (xhigh) leads on quality: 65.4 vs 63.9.
- GLM-5.2 (max) is stronger in agents & tools, composite, human preference, reasoning.
- GPT-5.4 (xhigh) is stronger in coding, instruction following, knowledge, long context, maths, multimodal.
- GLM-5.2 (max) is 2.6× cheaper ($2.15 vs $5.63 per 1M blended).
- GPT-5.4 (xhigh) streams 1.7× faster (121 vs 71 tokens per second).
| Metric | GLM-5.2 (max) | GPT-5.4 (xhigh) |
|---|---|---|
| BenchLeader Index | 63.9 | 65.4 |
| Agents & tools score | 63.8 | 63.4 |
| Coding score | 63.8 | 66.6 |
| Composite score | 72.1 | 69.4 |
| Human preference score | 67.2 | – |
| Instruction following score | 71.6 | 72.1 |
| Knowledge score | 55.5 | 60.6 |
| Long context score | 65.6 | 67.5 |
| Maths score | 59.0 | 66.4 |
| Reasoning score | 72.9 | 72.4 |
| Multimodal score | – | 62.9 |
| Blended price $/M | $2.15 | $5.63 |
| Output speed | 71 tok/s | 121 tok/s |
| Time to first answer | 31.6 s | 141.3 s |
| Context window | 1M | 1.1M |
| GPQA Diamond | 91.9% | 93.3% |
| FrontierMath Tiers 1–3 | 59.2% | 78.6% |
| FrontierMath Tier 4 | 29.3% | 49.0% |
| OTIS Mock AIME | 86.4% | 95.3% |
| SWE-bench Verified (Epoch) | 78.7% | – |
| SimpleQA Verified | 34.2% | 45.1% |
| Humanity's Last Exam | – | 36.2% |
| SciCode | 50.5% | 56.6% |
| WeirdML | 70.1% | 77.7% |
| APEX-Agents | – | 36.0% |
| ProofBench | 35.0% | 56.0% |
| GSO-Bench | – | 31.4% |
| LMArena Text | 1472 | – |
| LMArena Hard Prompts | 1493 | – |
| LMArena Coding | 1510 | – |
| LMArena WebDev | 1592 | – |
| LMArena Agent | 4.4 | – |
| LiveBench | – | 78.0% |
| LiveBench Reasoning | – | 88.1% |
| LiveBench Coding | – | 77.5% |
| LiveBench Agentic Coding | – | 53.8% |
| LiveBench Mathematics | – | 94.2% |
| LiveBench Data Analysis | – | 79.3% |
| LiveBench Language | – | 82.6% |
| LiveBench Instruction Following | – | 70.2% |
| AA Intelligence Index | 34.0 | 39.0 |
| IFBench | 73.3% | 74.0% |
| AA-LCR | 78.3% | 82.0% |
| MMMU-Pro | – | 78.4% |
| AA-Omniscience | 4.4 | 5.8 |
| Terminal-Bench Hard | 50.8% | 57.6% |
| GPQA Diamond (AA) | 89.5% | 92.0% |
| Humanity's Last Exam (AA) | 41.1% | 43.7% |
| SciCode (AA) | 51.2% | – |
| τ²-Bench Telecom (AA) | 99.1% | 87.1% |
| AIME (Vals) | – | 96.7% |
| LiveCodeBench | – | 84.1% |
| MMLU-Pro | – | 87.5% |
| LegalBench | – | 86.0% |
| CorpFin | – | 65.3% |
| TaxEval | – | 74.0% |
| MedQA | – | 96.1% |
| Terminal-Bench 2.1 (Vals) | 67.8% | – |
| SWE-bench (Vals) | 82.8% | 78.2% |
| GPQA Diamond (Vals) | – | 91.7% |
| AIME 2026 | – | 99.2% |
| HMMT February 2026 | – | 97.7% |
| MathArena Apex | – | 54.2% |
| SWE-Bench Pro | – | 59.1% |
| MCP Atlas | – | 70.6% |
| ARC-AGI-1 | – | 93.7% |
| ARC-AGI-2 | – | 74.0% |
| CritPt | 20.9% | 23.4% |
| GDPval (AA) | 45.3% | 40.3% |
| τ²-Bench Banking (AA) | 34.6% | 39.6% |
| ITBench SRE (AA) | 42.7% | – |
| APEX-Agents (AA) | 33.7% | 33.3% |
| CaseLaw v2 | – | 63.8% |
| Code Migration | 37.9% | 35.0% |
| Harvey's Legal Agent Benchmark | 7.1% | 0.0% |
| Legal Research Bench | 31.3% | – |
| MedCode | – | 41.3% |
| MedScribe | – | 77.5% |
| MMMU-Pro (Vals) | – | 87.5% |
| MortgageTax | – | 68.3% |
| ProgramBench | 0.5% | 0.0% |
| SAGE | – | 43.3% |
| SkillsBench | 45.1% | 51.7% |
| SREBench | 0.0% | – |
| Vibe Code Bench v1.1 | 64.0% | 67.4% |
| SWE-Bench Pro (private) | – | 43.4% |
| SWE Atlas: Codebase QnA | – | 40.8% |
| SWE Atlas: Refactoring | – | 44.3% |
| SWE Atlas: Test Writing | – | 44.4% |
| LMArena Maths | 1480 | – |
| LMArena Creative Writing | 1451 | – |
| LMArena Instruction Following | 1466 | – |
| LMArena Multi-turn | 1469 | – |
| LMArena Longer Queries | 1483 | – |
| Chess Puzzles | 21.0% | 44.0% |
| EBR-bench | 9.5% | 25.4% |
| Mystery Game Puzzles | – | 37.0% |
| PostTrainBench | 31.7% | – |
| CL-bench | – | 27.9% |
| CL-bench Life | – | 21.7% |
| METR Time Horizons | – | 74.3% |
| DeepSWE | 43.8% | 51.8% |
| LMCA | 45.8% | 52.0% |
| DTBench | 93.6% | 94.4% |
| 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.
GLM-5.2 vs GPT-5.4: questions
- Is GLM-5.2 better than GPT-5.4?
- GPT-5.4 (xhigh) leads on quality: 65.4 vs 63.9. The BenchLeader Index combines every independent quality benchmark; GPT-5.4 (xhigh) is ahead overall as of 2026-09-19, but check the category scores for your use.
- Is GLM-5.2 better than GPT-5.4 for coding?
- GPT-5.4 scores higher in coding (67 vs 64 on the category index, where 50 is average).
- Is GLM-5.2 better than GPT-5.4 for agentic tasks?
- GLM-5.2 scores higher in agentic tasks (64 vs 63 on the category index, where 50 is average).
- Which is cheaper, GLM-5.2 or GPT-5.4?
- GLM-5.2 is cheaper: $2.15 against $5.63 per million tokens, blended at three input tokens per output token.
- Which is faster, GLM-5.2 or GPT-5.4?
- GPT-5.4 streams faster: 121 against 71 output tokens per second.
- Which has the larger context window?
- GPT-5.4 accepts more context: 1.1M against 1M tokens.