Benchmarks · Reasoning & puzzles · Knowledge & accuracy
LisanBench
Word chains: from each of 50 starting words, the model links real words that each differ from the last by one letter (added, removed or swapped), never repeating a word, for as long as it can. Tests vocabulary, planning and rule-following. Valid steps per chain, summed over the 50 words (three tries each). By @scaling01.
116 models · 190 results · data as of 2026-11-16 · lisanbench.com ↗
Leaderboard
Muse Wonder Full xhigh
More from this benchmark
4 published measuresScore vs any published measure4 measures: sub-scores, splits, counts
104 more models under All models in settings.
Breakdown
darker is better within each column · click a column to sort| Model | |||
|---|---|---|---|
| Muse Wonder Full xhigh | 9,337 | 1 | 149.0 |
| Muse Wonder max | 8,303 | 0.999 | 150.0 |
| Opus 5.5 xhigh | 5,963 | 0.999 | 150.0 |
| DeepSeek V4.1 Flash 1017 max | 7,733 | 0.988 | 150.0 |
| Opus 5 high | 5,068 | 0.998 | 150.0 |
| Fable 5.1 xhigh | 4,573 | 0.995 | 150.0 |
| Opus 4.7 xhigh | 5,123 | 0.997 | 149.0 |
| Opus 4.6 | 3,526 | 0.988 | 150.0 |
| Sonnet 5.5 xhigh | 2,983 | 0.999 | 150.0 |
| Fable 5 medium | 4,562 | 0.998 | 150.0 |
| Haiku 5.5 xhigh | 2,711 | 0.997 | 150.0 |
| Sonnet 4.6 | 2,944 | 0.993 | 150.0 |
| Kimi K3 max | 3,526 | 0.997 | 150.0 |
| GPT-5.5 medium | 3,316 | 0.995 | 150.0 |
| Opus 4.8 high | 2,694 | 0.999 | 150.0 |
| GPT-6.1 Sol high | 4,004 | 0.98 | 150.0 |
| GPT-6 Sol high | 3,118 | 0.988 | 150.0 |
| Gemini 4 Argon high | 3,849 | 0.753 | 150.0 |
| GPT-6 Astra medium | 3,113 | 0.989 | 150.0 |
| Gemini 3.1 Pro high | 1,929 | 0.96 | 150.0 |
| Sonnet 5 high | 1,737 | 0.999 | 150.0 |
| Opus 4.5 | 2,204 | 0.947 | 150.0 |
| GPT-5.4 medium | 2,738 | 0.962 | 150.0 |
| GPT-5.6 Sol medium | 1,403 | 0.979 | 150.0 |
| Gemini 3 Pro high | 1,131 | 0.955 | 150.0 |
All results
190 rows| # | Model | Effort | LisanBench |
|---|---|---|---|
| 1 | Muse Wonder Full | xhigh | 32190.2 ±1358.2 |
| 2 | Muse Wonder | max | 27973.3 ±1312.4 |
| 3 | Muse Wonder Full | high | 24760.3 ±1207.3 |
| 4 | Opus 5.5 | xhigh | 24029.0 ±1316.5 |
| 5 | Muse Wonder | xhigh | 22588.7 ±1300.2 |
| 6 | DeepSeek V4.1 Flash 1017 | max | 21107.3 ±1290.0 |
| 7 | Muse Wonder Full | medium | 18245.3 ±795.3 |
| 8 | Muse Wonder | high | 18145.3 ±979.0 |
| 9 | Opus 5 | high | 15428.3 ±1044.4 |
| 10 | Fable 5.1 | xhigh | 15100.3 ±1557.4 |
| 11 | DeepSeek V4.1 Flash 1017 | high | 14750.0 ±968.4 |
| 12 | Opus 4.7 | xhigh | 14400.0 ±749.4 |
| 13 | Opus 4.6 | 14206.3 ±678.8 | |
| 14 | Muse Wonder | medium | 13807.0 ±754.9 |
| 15 | Sonnet 5.5 | xhigh | 13787.0 ±712.4 |
| 16 | Fable 5 | medium | 13680.3 ±1018.5 |
| 17 | Muse Wonder Full | low | 13338.3 ±794.9 |
| 18 | Haiku 5.5 | xhigh | 12897.0 ±1019.4 |
| 19 | Sonnet 4.6 | 12060.0 ±563.0 | |
| 20 | DeepSeek V4.1 Flash 1017 | low | 11638.3 ±866.2 |
| 21 | Kimi K3 | max | 10520.7 ±768.9 |
| 22 | GPT-5.5 | medium | 9844.7 ±854.5 |
| 23 | Opus 4.8 | high | 9463.0 ±766.3 |
| 24 | Muse Wonder | low | 8833.3 ±620.4 |
| 25 | GPT-6.1 Sol | high | 8736.0 ±1557.0 |
| 26 | Opus 5.5 | high | 8660.3 ±446.7 |
| 27 | GPT-6 Sol | high | 8222.0 ±1068.0 |
| 28 | Fable 5.1 | high | 8171.3 ±628.7 |
| 29 | Opus 5 | medium | 8112.7 ±651.7 |
| 30 | Haiku 5.5 | high | 7608.0 ±563.8 |