Radar / AI models / Inkling 2 Small weights

Inkling 2 Small weights: Thinking Machines' 594B MoE goes open under Apache 2.0

Thinking Machines Lab published the full weights of Inkling 2 Small on Hugging Face on 12 Nov: a 594B-parameter MoE with 17B active per token, image input, 1M-token context, and a 31 on the Artificial Analysis Intelligence Index, a week after its hosted debut on Tinker.

new launchmodels · open-weightsJEV traction 0.6added 2026-11-14
Open Inkling 2 Small weights →View on the radar

Why it matters

A frontier-scale open-weights model (594B total, 17B active) trained natively in 4-bit NVFP4 in just 68 hours, now downloadable under Apache 2.0 with API access across seven providers on day one.

What you could build with it

An indie developer could build a private document-analysis service that fine-tunes Inkling 2 Small on a company's own contracts and reports using Tinker's $1.50-per-million-token training, then serves it on rented Blackwell GPUs: a frontier-class in-house analyst without sending documents to a third-party API.

Does it hold up?

Too early to judge - the weights only landed on 12 Nov, the checkpoint needs roughly 500GB of aggregated VRAM, and no independent benchmarks beyond the vendor-reported Intelligence Index score of 31 have appeared yet.

Built with Inkling 2 Small weights

Learn more

technical deep dive →

First spotted on article: source.

More AI models

VoiceStudio: open-source, fully-local ElevenLabs alternative for voice cloning, dubbing, and audiobooksdebpalash/VoiceStudio is a desktop app that runs voice cloning, voice design, video dubbing, dictation, transcription…models · JEV 0.75Edge0-145B: open 145B MoE that runs on a single consumer GPUEdge0 open-sourced 145B and 120B latent-MoE models derived from Qwen3.8-Flash-Next, plus an inference framework that…models · JEV 0.74Strata: run a 125B model on a gaming PCFree, MIT-licensed local inference engine that runs Qwen3.8-Flash-Next (125B MoE, about 6B active per token) on…models · JEV 0.73Google AI Edge Gallery: offline speech recognition and synthesis on-deviceGoogle added fully offline speech recognition and text-to-speech to AI Edge Gallery, with ML Kit APIs on Android and…models · JEV 0.73Soup: fine-tune an 8B LLM on a 4 GB laptop GPU from one YAML fileAn open-source CLI that fine-tunes LLMs from a single YAML file, with layer streaming that trains an 8B model on a 4…models · JEV 0.73Tencent Hunyuan Hy3: open 295B MoE with 21B active under Apache 2.0Tencent open-sourced Hy3, a 295B-parameter mixture-of-experts model with only 21B parameters active per token, under…models · JEV 0.73

Get the week's best AI launches, plus 3 ideas worth building

One email every Saturday. Ranked by traction, not hype. Free.