Minos Opens macOS Beta for Subnet 107 Workload
The prototype app lets users draft evaluation assignments for a proposed Bittensor Subnet 107 workload, though it does not yet submit work on-chain.
Updated on Nov. 9, 2026 in Artificial Intelligence
Minos announced a public beta of its macOS app on November 6, 2026, offering a prototype for authoring evaluations on Bittensor Subnet 107. Beta participants receive no subnet rewards, and the workflow does not register or submit work on-chain.
Why it matters
The beta previews evaluation authoring as the proposed next workload for a subnet whose established work centers on genomic variant calling. It shows how a decentralized AI network might broaden the kinds of tasks it supports.
The beta requires an Apple Silicon Mac running macOS 14 or later, and users must bring their own AI subscription or credits from services such as Codex, Claude Code, Kimi Code, SayGM or Chutes. Beta participants receive no subnet rewards, and the workflow does not register or submit work on-chain, leaving the production timeline unannounced.
The players
Minos
Minos is the company that built the macOS app and prototype evaluation-authoring workflow for Bittensor Subnet 107.
Bittensor Subnet 107
Bittensor Subnet 107 is a decentralized network on the Bittensor protocol whose established work centers on genomic variant calling.
The details
Users' AI services draft a scientific brief from public literature for each assignment, and the Mac audits the draft before uploading an evidence bundle through a signed URL. Assignments are routed to Minos' platform test queue during the prototype phase.
Timeline
Friday, November 6, 2026, 11:50 p.m. ET (Saturday, November 7, 2026, 04:50 UTC): Minos announces the public beta of its macOS app.
Roadmap
Bittensor organizes decentralized AI work into specialized subnets, and Subnet 107 has so far focused on genomic variant calling. Minos' beta extends that model by previewing evaluation authoring as a second workload for the same network.
Developers with an Apple Silicon Mac can join the beta by supplying their own AI subscription or credits, but they will not earn subnet rewards during testing. Anyone interested in contributing to the prototype should expect the workflow to remain off-chain for now.
The takeaway
Decentralized AI networks are experimenting with new workloads before tying them to rewards. Readers following Bittensor should watch whether evaluation authoring eventually moves on-chain.
Further reading
For broader context on decentralized machine-intelligence networks, explore more in Artificial Intelligence.
Source note: This article includes information reported by TokenPost.
