Model comparison
Gemini 3 Pro vs GLM-5.2
Gemini 3 Pro is the stronger model overall, scoring 54.8 to 51.1 on the Noometry Index.
Last verified . 36 shared benchmarks.
GoogleSummary
- They share 36 benchmarks with published results for both. Gemini 3 Pro scores higher in 5 categories and GLM-5.2 in 4 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Gemini 3 Pro leads 52.5 to 42.3.
- The biggest single-benchmark swing is NYT Connections (extended): 94.4% for Gemini 3 Pro and 74.3% for GLM-5.2.
- GLM-5.2 has downloadable open weights; the other is API-only.
Side by side
| Gemini 3 Pro | GLM-5.2 | |
|---|---|---|
| Provider | Z.ai (Zhipu) | |
| Noometry Index | 54.8 | 51.1 |
| Released | 2025-11-18 | 2026-06-13 |
| Weights | Proprietary | Open |
| Context window | — | 1M |
| Max output | — | 131K |
| Input $ / M tokens | — | $1.40 |
| Output $ / M tokens | — | $4.40 |
| Results tracked | 67 | 51 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Gemini 3 Pro: 51.6 (#39), GLM-5.2: 51.3 (#41)
| Benchmark | Gemini 3 Pro | GLM-5.2 |
|---|---|---|
| SWE-bench Verified | 72.9% | 78.7% |
| LMArena WebDev | 1440 | 1603 |
| WeirdML | 69.9% | 70.1% |
| LMArena Coding | 1481 | 1485 |
| ALE-Bench | 1,177 | 1,047 |
| DeepSWE | — | 43.8% |
| FrontierCode | — | 24.5% |
| SWE-bench Verified (bash only) | 74.2% | — |
| SWE-bench Multilingual | 68.7% | — |
| SciCode | — | 50.5% |
| GSO | 18.6% | — |
| AlgoTune | 1.83 | — |
Agentic & Tool Use Gemini 3 Pro leads
Gemini 3 Pro: 40.6 (#23), GLM-5.2: 32.4 (#63)
| Benchmark | Gemini 3 Pro | GLM-5.2 |
|---|---|---|
| τ²-bench Banking | 18% | 37.1% |
| Vending-Bench 2 | 5,478 | 8,314 |
| Terminal-Bench | 69.4% | — |
| APEX-Agents | — | 45.2% |
| Berkeley Function Calling Leaderboard | 72.5% | — |
| GDPval | 40.3% | — |
| Remote Labor Index | 1.3% | — |
| τ²-bench Airline | 80.5% | — |
| τ²-bench Retail | 75.9% | — |
| τ²-bench Telecom | 91% | — |
| DeepResearch Bench | 46.3% | — |
| PostTrainBench | — | 31.7% |
| BALROG | 58.1% | — |
| GBAEval | — | 0% |
| LMArena Search | 1207 | — |
| METR Time Horizons | 71% | — |
Reasoning Gemini 3 Pro leads
Gemini 3 Pro: 52.5 (#31), GLM-5.2: 42.3 (#52)
| Benchmark | Gemini 3 Pro | GLM-5.2 |
|---|---|---|
| ARC-AGI-2 | 31.1% | 22.8% |
| SimpleBench | 76.4% | 58.8% |
| Kagi LLM Benchmark | 80.1% | 62.6% |
| NYT Connections (extended) | 94.4% | 74.3% |
| ARC-AGI-1 | 75% | 77% |
| CritPt | 6.9% | 20.9% |
| Chess Puzzles | 31% | 21% |
| LMArena Hard Prompts | 1480 | 1480 |
| Epoch Capabilities Index | 152.92 | 151.78 |
| EnigmaEval | 18.2% | — |
| EBR-Bench | — | 9.5% |
| Mystery Game Puzzles | — | 19% |
| DTBench | — | 93.6% |
| LMCA | — | 45.8% |
| Surface Evolver Bench | — | 55.6% |
| ForecastBench | 61.2 | — |
Math GLM-5.2 leads
Gemini 3 Pro: 49.9 (#59), GLM-5.2: 55.7 (#43)
| Benchmark | Gemini 3 Pro | GLM-5.2 |
|---|---|---|
| MathArena Final-Answer Competitions | 67% | 67.6% |
| OTIS Mock AIME 2024-2025 | 91.4% | 86.4% |
| ProofBench | 20% | 35% |
| LMArena Math | 1476 | 1482 |
| FrontierMath (Tiers 1-3) | — | 59.2% |
| FrontierMath Tier 4 | — | 29.3% |
| Omni-MATH | 55.5% | — |
| FrontierMath (Feb 2025 set) | 37.6% | — |
| FrontierMath Tier 4 (v1) | 18.8% | — |
Knowledge Gemini 3 Pro leads
Gemini 3 Pro: 64.4 (#16), GLM-5.2: 57.1 (#40)
| Benchmark | Gemini 3 Pro | GLM-5.2 |
|---|---|---|
| GPQA Diamond | 92.6% | 91.9% |
| LMArena Expert | 1475 | 1486 |
| Humanity's Last Exam | 37.5% | — |
| SimpleQA Verified | — | 34.2% |
| MMLU-Pro | 90.3% | — |
| Vectara Hallucination Rate | 13.6% | — |
| GPQA (HELM) | 80.3% | — |
Multimodal Not comparable
Gemini 3 Pro: 57.6 (#2), GLM-5.2: —
| Benchmark | Gemini 3 Pro | GLM-5.2 |
|---|---|---|
| LMArena Vision | 1305 | — |
| GeoBench | 84% | — |
| VPCT | 91% | — |
| LMArena Document | 1434 | — |
Multilingual Gemini 3 Pro leads
Gemini 3 Pro: 56.9 (#16), GLM-5.2: 55.8 (#26)
| Benchmark | Gemini 3 Pro | GLM-5.2 |
|---|---|---|
| LMArena Non-English | 1474 | 1459 |
| LMArena Chinese | 1523 | 1519 |
| LMArena French | 1492 | 1479 |
| LMArena German | 1515 | 1468 |
| LMArena Japanese | 1510 | 1451 |
| LMArena Korean | 1448 | 1445 |
| LMArena Russian | 1493 | 1466 |
| LMArena Spanish | 1470 | 1477 |
Instruction Following Too close to call
Gemini 3 Pro: 76.3 (#45), GLM-5.2: 76.9 (#34)
| Benchmark | Gemini 3 Pro | GLM-5.2 |
|---|---|---|
| LMArena Instruction Following | 1458 | 1465 |
| IFEval | 87.7% | — |
Long Context GLM-5.2 leads
Gemini 3 Pro: 44.0 (#79), GLM-5.2: 45.3 (#43)
| Benchmark | Gemini 3 Pro | GLM-5.2 |
|---|---|---|
| LMArena Longer Query | 1471 | 1479 |
| CL-bench | 15.8% | — |
Writing & Preference GLM-5.2 leads
Gemini 3 Pro: 66.4 (#35), GLM-5.2: 70.4 (#21)
| Benchmark | Gemini 3 Pro | GLM-5.2 |
|---|---|---|
| LMArena Text | 1479 | 1470 |
| LMArena Creative Writing | 1482 | 1462 |
| EQ-Bench Creative Writing | 1525 | 1757 |
| LMArena Multi-Turn | 1484 | 1469 |
| WildBench | 85.9% | — |
| EQ-Bench 4 | — | 1222 |
Frequently asked questions
Is Gemini 3 Pro better than GLM-5.2?
Gemini 3 Pro is the stronger model overall, scoring 54.8 to 51.1 on the Noometry Index.
Is Gemini 3 Pro or GLM-5.2 better for coding?
They score almost the same on coding (51.6 vs 51.3); test both on your own repository before choosing.
How many benchmarks do Gemini 3 Pro and GLM-5.2 share?
36 benchmarks have published results for both models. Gemini 3 Pro has 67 scored results on Noometry and GLM-5.2 has 51.