Compare / Kimi K2 Thinking vs GLM 5.1
Kimi K2 Thinking vs GLM 5.1
16 results from 3 sources that measured both models. Each row is in the source's own units; there is no overall winner.
| Benchmark | Metric | Kimi K2 Thinking | GLM 5.1 |
|---|---|---|---|
| Artificial Analysisreported | Artificial Analysis Intelligence Indexindex score | 22.0 | 26.1 |
| Artificial Analysisreported | GPQA Diamond% of questions | 83.8% | 86.8% |
| Artificial Analysisreported | Humanity's Last Exam% of questions | 23.8% | 30.1% |
| Artificial Analysisreported | IFBench% of instructions | 68.1% | 76.3% |
| Artificial Analysisreported | Long-context reasoning (AA-LCR)% of questions | 72.0% | 73.7% |
| Artificial Analysisreported | Terminal-Bench Hard% of tasks | 31.1% | 43.2% |
| Artificial Analysisreported | Τ²-bench telecom% of tasks | 93.0% | 97.7% |
| OpenHands Indexreported | Average% resolved | 41.0% | 58.2% |
| OpenHands Indexreported | Front end (SWE-Bench Multimodal)% resolved | 32.4% | 41.2% |
| OpenHands Indexreported | Greenfield (Commit0)% resolved | 12.5% | 37.5% |
| OpenHands Indexreported | Information gathering (GAIA)% resolved | 43.6% | 67.3% |
| OpenHands Indexreported | Issue resolution (SWE-Bench)% resolved | 69.2% | 75.0% |
| OpenHands Indexreported | Testing (SWT-Bench)% resolved | 47.3% | 70.2% |
| UGI Leaderboardreported | Requested-length error% off the requested word count | 20.0% | 9.0% |
| UGI Leaderboardreported | Style adherencescore from 0 to 1 | 0.38 | 0.38 |
| UGI Leaderboardreported | Writing scorescore out of 100 | 59.3 | 61.7 |
Bold green marks the better value on that metric. Where a source reports ranges that overlap, the difference may not be meaningful; see the benchmark page for ranges.