Comparing 2 models
Kimi K3 vs GLM 5.2
Benchmarks through
Verdict
Kimi K3 places higher on 7 of the 10 tasks where both are ranked. GLM 5.2 costs 7.0x less per input token.
- Choose Kimi K3 for Local LLM for Coding (#1), Structured Data Extraction (#3), JSON & Schema Output (#3).
- Choose GLM 5.2 for Cheapest Capable LLM (#2), Cheap LLM for High Volume (#2), High-Volume Extraction (#2).
Built from the published task rankings below; ranks come from each task's own weighted benchmarks, so a win on one task says nothing about another.
Measured head to head
Benchmarks both models were run on
| Evaluation | Kimi K3 | GLM 5.2 | Source |
|---|---|---|---|
| Design Arena Coding | 1360 Elo | 1287 Elo | source · 2026-10-23 |
| Design Arena Full-stack Agent | 1315 Elo | 1241 Elo | source · 2026-10-22 |
| Design Arena UI | 1341 Elo | 1294 Elo | source · 2026-10-23 |
| Design Arena Web-app Agent | 1295 Elo | 1222 Elo | source · 2026-10-23 |
| Design Arena Website | 1335 Elo | 1285 Elo | source · 2026-10-23 |
| FrontierMath Tiers 1-3 (Epoch AI) | 72.2% ± 2.7 | 59.2% ± 3 | source · 2026-07-17 |
| GPQA Diamond (Epoch AI) | 91.9% ± 1.9 | 91.9% ± 1.6 | source · 2026-08-07 |
| GPQA Diamond (Epoch AI) | 93.1% ± 1.5 | 91.9% ± 1.6 | source · 2026-07-16 |
| LiveBench Agentic Coding | 62.2% | 51.8% | source · 2026-10-18 |
| LiveBench Coding | 81.5% | 79.7% | source · 2026-10-18 |
| LiveBench Data Analysis | 78.7% | 73.7% | source · 2026-10-18 |
| LiveBench Instruction Following | 71.4% | 62.3% | source · 2026-10-18 |
| LiveBench Language | 85.5% | 76.2% | source · 2026-10-18 |
| LiveBench Math | 84.4% | 89.8% | source · 2026-10-18 |
| LiveBench Reasoning | 90.7% | 78.6% | source · 2026-10-18 |
| LMArena Agent | 1.8 | 2.1 | source · 2026-10-20 |
| LMArena Agent Bash Recovery | 4.4 | 0.7 | source · 2026-10-08 |
| LMArena Agent Steerability | 0.4 | 1.3 | source · 2026-10-20 |
| LMArena Agent Task Outcome | 6.3 | 3.8 | source · 2026-10-20 |
| LMArena Agent Tool Hallucination | 0.7 | 0.7 | source · 2026-10-20 |
| LMArena Creative Writing | 1456 Elo | 1462 Elo | source · 2026-10-22 |
| LMArena Instruction Following | 1485 Elo | 1466 Elo | source · 2026-10-22 |
| LMArena Long Query | 1493 Elo | 1479 Elo | source · 2026-10-22 |
| LMArena Math | 1493 Elo | 1484 Elo | source · 2026-10-22 |
| LMArena Text | 1487 Elo | 1475 Elo | source · 2026-10-22 |
| LMArena Text Factuality | 1472 Elo | 1463 Elo | source · 2026-10-22 |
| LMArena WebDev | 1655 Elo | 1604 Elo | source · 2026-10-22 |
| OTIS Mock AIME (Epoch AI) | 93.3% ± 3.8 | 86.4% ± 4.2 | source · 2026-08-07 |
| OTIS Mock AIME (Epoch AI) | 97.2% ± 1.1 | 86.4% ± 4.2 | source · 2026-07-16 |
| SimpleQA Verified (Epoch AI) | 50.6% ± 1.6 | 34.2% ± 1.5 | source · 2026-08-27 |
| tau2-bench (1.0.1) | 37.1% | 37.1% | source · 2026-07-24 |
Values are as published by each evaluator, in its own units; rows from different evaluators are not on a common scale.
task evidence
How they place when the question changes
Lower ranks are better. “Best” marks the strongest option in this shortlist for that task.