Gemini 3.1 Pro vs GLM-5.2
Gemini 3.1 Pro 
GLM-5.2 Verdict
- Gemini 3.1 Pro and GLM-5.2 (max) are level on quality (63.9 vs 63.9).
- Gemini 3.1 Pro is stronger in knowledge, long context, maths, multimodal.
- GLM-5.2 (max) is stronger in agents & tools, coding, composite, human preference, instruction following, reasoning.
- GLM-5.2 (max) is 2.1× cheaper ($2.15 vs $4.50 per 1M blended).
- Gemini 3.1 Pro streams 1.7× faster (121 vs 71 tokens per second).
| Metric | Gemini 3.1 Pro | GLM-5.2 (max) |
|---|---|---|
| BenchLeader Index | 63.9 | 63.9 |
| Agents & tools score | 60.4 | 63.8 |
| Coding score | 59.8 | 63.8 |
| Composite score | 67.5 | 72.1 |
| Human preference score | 59.8 | 67.2 |
| Instruction following score | 69.7 | 71.6 |
| Knowledge score | 65.9 | 55.5 |
| Long context score | 67.5 | 65.6 |
| Maths score | 61.9 | 59.0 |
| Multimodal score | 66.0 | – |
| Reasoning score | 70.8 | 72.9 |
| Blended price $/M | $4.50 | $2.15 |
| Output speed | 121 tok/s | 71 tok/s |
| Time to first answer | 34.8 s | 31.6 s |
| Context window | 1.0M | 1M |
| GPQA Diamond | 94.1% | 91.9% |
| FrontierMath Tiers 1–3 | 59.6% | 59.2% |
| FrontierMath Tier 4 | 26.8% | 29.3% |
| OTIS Mock AIME | 95.6% | 86.4% |
| SWE-bench Verified (Epoch) | – | 78.7% |
| SimpleQA Verified | – | 34.2% |
| Humanity's Last Exam | 46.4% | – |
| Terminal-Bench | 80.2% | – |
| SimpleBench | 79.6% | – |
| SciCode | 58.9% | 50.5% |
| WeirdML | 72.1% | 70.1% |
| APEX-Agents | 33.5% | – |
| ProofBench | 26.0% | 35.0% |
| GSO-Bench | 22.6% | – |
| Epoch Capabilities Index | 155 | – |
| LMArena Text | 1487 | 1472 |
| LMArena Hard Prompts | 1508 | 1493 |
| LMArena Coding | 1520 | 1510 |
| LMArena WebDev | 1447 | 1592 |
| LMArena Vision | 1296 | – |
| LMArena Agent | -5.8 | 4.4 |
| AA Intelligence Index | 30.4 | 34.0 |
| IFBench | 77.1% | 73.3% |
| AA-LCR | 82.0% | 78.3% |
| MMMU-Pro | 82.4% | – |
| AA-Omniscience | 31.9 | 4.4 |
| Terminal-Bench Hard | 53.8% | 50.8% |
| GPQA Diamond (AA) | 94.1% | 89.5% |
| Humanity's Last Exam (AA) | 47.0% | 41.1% |
| SciCode (AA) | 58.7% | 51.2% |
| τ²-Bench Telecom (AA) | 95.6% | 99.1% |
| Terminal-Bench 2.1 (Vals) | – | 67.8% |
| SWE-bench (Vals) | – | 82.8% |
| AIME 2026 | 98.3% | – |
| HMMT February 2026 | 94.7% | – |
| MathArena Apex | 60.9% | – |
| MultiChallenge | 71.4% | – |
| PRBench Finance | 41.9% | – |
| PRBench Legal | 44.0% | – |
| MultiNRC | 64.7% | – |
| HiL-Bench | 35.3% | – |
| TutorBench | 53.0% | – |
| EQ-Bench 4 | 1142 | – |
| ARC-AGI-1 | 98.0% | – |
| ARC-AGI-2 | 77.1% | – |
| ARC-AGI-3 | 0.4% | – |
| CritPt | 17.7% | 20.9% |
| GDPval (AA) | 20.2% | 45.3% |
| τ²-Bench Banking (AA) | 21.4% | 34.6% |
| ITBench SRE (AA) | 30.3% | 42.7% |
| Analyst Agent (AA) | 41.3% | – |
| APEX-Agents (AA) | 32.0% | 33.7% |
| Code Migration | – | 37.9% |
| CyberBench | 36.4% | – |
| Harvey's Legal Agent Benchmark | – | 7.1% |
| Legal Research Bench | – | 31.3% |
| ProgramBench | – | 0.5% |
| SkillsBench | – | 45.1% |
| SREBench | – | 0.0% |
| Vibe Code Bench v1.1 | – | 64.0% |
| FORTRESS | 29.8% | – |
| MASK | 42.4% | – |
| SWE Atlas: Codebase QnA | 13.5% | – |
| SWE Atlas: Refactoring | 33.8% | – |
| SWE Atlas: Test Writing | 29.8% | – |
| VTB | 29.0% | – |
| LMArena Maths | 1489 | 1480 |
| LMArena Creative Writing | 1480 | 1451 |
| LMArena Instruction Following | 1481 | 1466 |
| LMArena Multi-turn | 1495 | 1469 |
| LMArena Longer Queries | 1500 | 1483 |
| LMArena Document | 1459 | – |
| Chess Puzzles | 55.0% | 21.0% |
| EBR-bench | 14.3% | 9.5% |
| BALROG | 57.0% | – |
| PostTrainBench | 22.0% | 31.7% |
| ExploitBench | 26.1% | – |
| CL-bench | 20.8% | – |
| CL-bench Life | 16.9% | – |
| METR Time Horizons | 77.0% | – |
| DeepSWE | 11.7% | 43.8% |
| LMCA | – | 45.8% |
| DTBench | – | 93.6% |
| ForecastBench | 59.0% | – |
| CursorBench | – | 55.0% |
| GBAEval | 0.8% | – |
| ALE-Bench | 1160.6 | 1010.2 |
| AlgoTune | 2.0 | – |
| Vending-Bench 2 | 911.2 | – |
| Blueprint-Bench 2 | 26.5% | – |
| GDP.pdf | 17.0% | – |
Data as of 2026-09-19. Best configuration of each model; every score links to its source on the model pages.
Gemini 3.1 Pro vs GLM-5.2: questions
- Is Gemini 3.1 Pro better than GLM-5.2?
- Gemini 3.1 Pro and GLM-5.2 (max) are level on quality (63.9 vs 63.9). The BenchLeader Index combines every independent quality benchmark; GLM-5.2 (max) is ahead overall as of 2026-09-19, but check the category scores for your use.
- Is Gemini 3.1 Pro better than GLM-5.2 for coding?
- GLM-5.2 scores higher in coding (64 vs 60 on the category index, where 50 is average).
- Is Gemini 3.1 Pro better than GLM-5.2 for agentic tasks?
- GLM-5.2 scores higher in agentic tasks (64 vs 60 on the category index, where 50 is average).
- Which is cheaper, Gemini 3.1 Pro or GLM-5.2?
- GLM-5.2 is cheaper: $2.15 against $4.50 per million tokens, blended at three input tokens per output token.
- Which is faster, Gemini 3.1 Pro or GLM-5.2?
- Gemini 3.1 Pro streams faster: 121 against 71 output tokens per second.
- Which has the larger context window?
- Gemini 3.1 Pro accepts more context: 1.0M against 1M tokens.