The answer to agents uploading your code is an agent that can't. jetbrains' junie local runs a 27b…
the answer to agents uploading your code is an agent that can't. jetbrains' junie local runs a 27b qwen model fully on-device — no cloud, no tokens, no registration — and the cheapest m5 mac with 64gb of ram that runs it is about $2,700.
on-device sovereignty has a hardware price.
Context
JetBrains' launch post of 24 August 2026 says Junie Local runs Qwen3.6-27B at 4-bit, about a 20 GB download, on an M5 Mac with 64 GB of RAM, with no tokens, no quota, no registration, no card, and works without a network after the download. The docs say it runs through a managed mlx-vlm server on localhost, needs Apple Silicon M5 or newer and about 21 GB of disk, and was in nightly builds when read. JetBrains reports quality on its own private test set as on par with Sonnet 4.5 with a 10,000-token reasoning limit, with GPT-5 at medium effort slightly higher and the local model's reasoning disabled. The New Stack of 24 August 2026 calls 64 GB firmly a high-end Mac proposition.
The model, size, hardware floor and free access are first-party. The 2,700 dollar figure is not on any page read, including The New Stack; it needs a dated Apple price source and is unverified. The code stays on the machine, with the one-time model download as the only network step. The quality comparison is the vendor's, on a private set. That on-device sovereignty has a hardware price is the author's framing.
Watch next
- Apple's current price for the cheapest 64 GB M5 configuration, and whether the feature is in stable builds.
Sources
- Junie Local launch (JetBrains, 24 Aug 2026)blog.jetbrains.com
- Junie Local docs (JetBrains)junie.jetbrains.com
- JetBrains Junie Local agent (The New Stack, 24 Aug 2026)thenewstack.io
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 20 September 2026 at 03:43 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →