Inkling 2 Mini first impressions: cracks spatial puzzle no local model could solve
MerePotato · reddit · 2026-11-23
Testing Inkling 2 Mini's official mixed-precision release on a private general-reasoning benchmark (100 novel puzzles, riddles and physics questions), it scored 70% — matching but not surpassing Qwen 3.8 27B. Notably, it solved a spatial reasoning puzzle that no other local model the author has tested could crack.
More from Models
- Replit agents speed up NumPy 18.9x — but only Claude's team scales — amasad · 2026-11-23
- Perplexity is a flawed cross-lab capability bound: tokenizer granularity favors OpenAI — davidmanheim · 2026-11-23
- Gemini 4 Hydrogen at $1.50/M Input Tokens Scores 51, Beating Argon's 50 — dl_weekly · 2026-11-23
- Base log perplexity is an extremely strong predictor of Epoch and OpenAI ECI, researchers note — stochasticchasm · 2026-11-23
- Redis creator antirez: AI is already great at math and programming, blaming the tool is the real problem — antirez · 2026-11-23
- Gemini 4 xhigh costs ~2.8x more but passes the same 12 checks as high in coding test — mhdfaran · 2026-11-23