Inkling-2-Mini on 16GB VRAM: 98% in 1h05m but 'not 27B tier' in real tasks
Distinct-Pie2389 · reddit · 2026-11-23
A 16GB VRAM user benchmarked Inkling-2-Mini locally (Q4KXL, 262K ctx, INT8 KV): on the same airbench suite it scored 98% (best 2/3) in 1 hour vs 1h24m for the 27B model.
Verdict: faster, but 'cheeks' on real tasks and behind even on standard benchmarks — nowhere near 27B tier. Only worth it for machines that can't fit 27B, where the Q4→Q1 quant degradation curve should be gentler.
More from Models
- OpenAI chief scientist: losses below 0.8 unlocked emergent cybersecurity and math capabilities — kimmonismus · 2026-11-23
- Dev baffled by Google's new Neon, Argon, Carbon, Hydrogen model codenames — eigenhector · 2026-11-23
- Gemini Pro usage screenshot shows Claude/GPT weekly quota fully exhausted at 0% — MediocreInside8628 · 2026-11-23
- AI agents play a 3D murder mystery: GPT-6.2 solves it in 3m14s — venturetwins · 2026-11-23
- Users say GPT-5.6 was the last model 'with the sauce' as newer ones feel robotic — thederbiedone · 2026-11-23
- Pretraining perplexity alone can't rank base models anymore — stochasticchasm · 2026-11-23