Search

Across 58925 posts, found59posts across0collections

All Collections 10 Posts 59

SortTitle matches first, then newest

Matching posts

Matching terms are highlighted

Liquid AI releases LFM3 on-device foundation model family

…dation models yet, built for specialization and on-device deployment. Here’s what’s new: > More capabilit…

模型发布@liquidai　Filed in No.004

23 hours ago

Grok Bot gains on-device app control on iOS and Android

Grok Bot now reads and sends messages, manages calendar, reminders and contacts, finds nearby places and controls smart home devices, so users can ask it to finish everyday tasks instead of switching apps.

personal-agent@XFreeze　Filed in No.012

5 days ago

PrismML releases Bonsai 2 on-device ternary models

…d supporting reasoning, vision and tool calling on-device.

模型发布@pratikg　Filed in No.004

12 days ago

Underdog launches Stripe Shop with on-device agent payments

…t amount, and waits for your approval The first on-device Agent Payment in Underdog: Your Private Persona…

agent@UnderdogAI　Filed in No.004

15 days ago

Developer makes Grok Bot character color and shape adjustable on-device

A Japanese developer made the Grok Bot character's color and shape adjustable on the device side and showed how close the modified look comes to the original.

grok@AILogDev　Filed in No.012

16 days ago

Video asks why Google trails Meta Muse in on-device assistants

…gle has the phones, TPUs, models and people for on-device agents, but lacks the courage to ship a winning…

personal-agent@semidoped　Filed in No.012

23 days ago

POST

Apple on-device Foundation Model writes text while Jev picks options

…oundation Model + @typesafeai Jev - fm: Apple's on-device Foundation Model (OS 27). Free, offline, no API…

jev@jasonkneen　Filed in No.006

2026.09.28

POST

Google's new AI memory architecture keeps decryption keys on-device

Google's new memory architecture stores encrypted memories in the cloud with keys on the user's device, decrypting briefly in an isolated environment, while open questions remain about lost phones, device changes, and true forgetting.

ai@frxiaobei　Loose post

2026.09.24

Jev routes on-device models to turn voice memos into to-dos

…s in one call within milliseconds to pick which on-device model to run, with no LLM in the loop.

jev@desertantlabs　Filed in No.006

2026.09.22

Laya runs millisecond decisions on-device, 50x faster than Jev

After watching a demo, the author says Laya has only about 0.4B parameters and uses roughly 1GB of memory during inference, making millisecond-level decisions locally for high-frequency production use cases.

jev@moc_tx89　Filed in No.006

2026.09.21

Showing10of59posts

10 per batch

Keep scrolling to reveal posts 11–20