Inkling-Small: the quarter-size open model that out-codes its trillion-parameter sibling
Thinking Machines released Inkling-Small on July 30: an open-weights model with 276 billion total but only 12 billion active parameters. Despite being roughly a quarter the size of the company's 975-billion-parameter Inkling, it scores higher on coding benchmarks — 80.2% versus 77.6% on SWE-bench Verified — while costing around $0.30/$1.20 per million tokens. The company concedes the smaller model knows fewer facts and trails on some agentic tasks.
Why it mattersFlagship-level coding skill now fits in a model cheap enough to run widely, and the weights are freely available on Hugging Face. For developers this keeps pushing the price of serious coding assistance down; for closed-model labs, the open-weights floor just rose again.
Read in the app — free, in 9 languages
Related stories
Rumour: the anonymous model that just topped a coding benchmark, for free, is said to be Zhipu's unreleased flagship2026-08-21DeepSeek's cheap workhorse can now see — and on agent tasks that need eyes it says it is close to Anthropic's best
2026-08-21Nvidia is paying $6 billion for the machine that builds a rival's models — and hiring 109 of the people who ran it
2026-08-21Given four hours and a GPU to improve the way AI is trained, the best agent scored 0.25 out of 1 — and most never tried
2026-08-21Show the robot once — three to twelve seconds — and it gets the job right 59 times out of 100 with no training at all
2026-08-21