✳ the wire · intel
New on the radar: Inkling 2 Small appeared in models.dev (opencode)
An email when every new Thinking Machines model answers on a public API, about 15 minutes after members, plus every frontier launch · no card
Our hourly sweep found an id that was not there an hour ago.
Catalog id: opencode/inkling-2-small-free
Listed price: $0/$0 per 1M
Context window: 1,048,576 tokens
First observed by our sweep: 2026-11-07 01:39 UTC
Source: models.dev (opencode)
This entry is machine-written from what the sweep actually read, and nothing else. A catalog listing is a primary source for exactly one narrow fact — that this id now exists, at this price, with this context window — and it is not evidence of anything beyond that. It does not tell you the model is good, that the listing is intentional, that the weights are open, or that the vendor has announced it. Those questions need a human and an editor's judgement, and this item deliberately does not answer them.
It maps to Inkling 2 Small, already on our watchlist — registry note: live — Nov 5 · alert fired by hand at 00:31:51 UTC Nov 6 (event aw_mw48oyj3_o21wrwdnj7, 19 notifications) because no catalog we sweep listed it · launched Nov 5 on Thinking Machines' Tinker only (its post 'Introducing Inkling 2 Small'): Tinker Playground, plus fine-tuning and sampling on Tinker, with closed weights for now; Thinking Machines says full weights follow next week under Apache 2.0, after external safety testing and red-teaming · 594B total / 17B active MoE, image input, variable effort · Thinking Machines says it scores 31 on the Artificial Analysis Intelligence Index · its own table: FrontierCode v2.0 25.4%, DeepSWE v1.1 72.1%, SWEBench Pro v2 (Hard) 86.3% · early checkpoints ran on OpenCode as Fledge-Alpha (owner statement in the launch post), which closes the fledge-alpha stealth slot onto this row · no inkling-2 id in OpenRouter or models.dev at 00:30 UTC Nov 6 · Nov 6: Artificial Analysis lists it at 31 on its Intelligence Index (#4 of 29 comparable models), $0.30 / $1.00 per 1M, 1M context, ~150 tok/s; Vals AI scores it 56.1 (first-gen Inkling: 29.0).
The radar reads the public model catalogs every hour and the provider APIs every 60 seconds. Signal texts you the minute a launch is confirmed; free alerts follow by email about 15 minutes later.

