Inkling 2 Mini quantized to 5.4GB/8.9GB, keeping 91-97% performance
teortaxesTex · x · 2026-11-21
Inkling 2 Mini — the open-weights model leading every sub-100B-total-parameter model on benchmarks like DeepSWE v1.1 (48.2%) and Vals Index v2.2 (46.5%), with 5.4B effective activations per token — now has quantized builds: shinboson released 5.4GB and 8.9GB versions retaining 91.2% and 96.8% of original performance respectively. A faster-drafter version and an iPhone version are coming soon. Good news for RAM-constrained local deployment.
More from Infra
- Open-weight models are less efficient, so their rise may increase chip sales — aronchick · 2026-11-21
- Score centering, rumored OpenAI RL trick, tolerates staleness and may rescue slow chips — teortaxesTex · 2026-11-21
- $31/month quad-channel EPYC + 128GB runs 3-bit Qwen at 128k context — Ne00n · 2026-11-21
- Ternary Bonsai-2 vs Qwen3.5-2B on A13/A14/M3: 32-55% faster decode, and why thermal logging matters — leo-k7v · 2026-11-21
- Essay 'The Organic Web' weaves 62 citations to dissect data's physical weight — round · 2026-11-21
- ARK projects Starship marginal launch cost falling to $358/kg by 2027 — mitchdeg · 2026-11-21