Global MMLU Lite — Inkling Release Comparisons, Unspecified Languages
General QA
Global MMLU Lite accuracy reported in the Inkling and Inkling-Small launch comparison tables. The language roster and exact question cohort are not established in the reviewed release documentation; these results remain diagnostic rather than being pooled with Artificial Analysis' verified 16-language aggregate.
Top models (higher is better)
| Model | Score |
|---|---|
| Claude Fable 5 | 93.3 |
| GPT-5.6 Sol | 91.8 |
| Gemini 3.5 Flash-Lite | 89.4 |
| DeepSeek-V4-Pro | 89.3 |
| GLM-5.2 | 89.2 |
| GPT-5.6 Luna | 88.7 |
| Inkling | 88.7 |
| DeepSeek-V4-Flash | 88.4 |
| Kimi K2.6 | 88.4 |
| Inkling-Small Preview | 86.8 |
| Inkling-Small | 86.7 |
| Nemotron 3 Ultra 550B A55B | 85.6 |
| MiniMax M2.7 | 83.9 |
| MiMo-V2.5 | 83.5 |