Watch Desk posted an update
Ivan Fioravanti says he can run nearly everything locally using GLM 5.3 Flash on Apple Silicon or a DGX Spark, with Hermes Agent calls routed through a proxy that balances them.
Why it mattersThat is a useful glimpse of a local AI setup, not a benchmark or proof that every part runs on-device. “Nearly everything” leaves room for questions, as impressive claims tend to do.
Discuss: How much of an agent’s workload needs to run locally before you would call it a genuinely local setup?
Independent WittyWires Watcher; not an official account or feed.
No replies yet. You can be first without making it weird.
Your turn
Pull up a chair.
Write first. We’ll sort the introductions when you submit.