LiquidAI_ Profile Banner
Liquid AI Profile
Liquid AI

@LiquidAI_

Followers
13K
Following
242
Media
46
Statuses
184

Build efficient general-purpose AI at every scale.

Cambridge, MA
Joined March 2023
Don't wanna be here? Send us removal request.
@LiquidAI_
Liquid AI
16 days
Today, we release the 2nd generation of our Liquid foundation models, LFM2. LFM2 set the bar for quality, speed, and memory efficiency in on-device AI. Built for edge devices like phones, laptops, AI PCs, cars, wearables, satellites, and robots, LFM2 delivers the fastest
Tweet media one
14
59
255
@LiquidAI_
Liquid AI
23 hours
RT @dottxtai: Edge inference demands blazing-fast structured generation. Our technology delivers exactly that and with grammars we can eli….
0
9
0
@LiquidAI_
Liquid AI
1 day
RT @LiquidAI_: We’ve partnered with @dottxtai and @remilouf to make function calling viable on the edge. JSON doesn’t cut it. We need stru….
0
15
0
@LiquidAI_
Liquid AI
1 day
RT @maximelabonne: LFM2 🤝 Outlines. Constrained generation is amazing to squeeze more performance from tiny models. We just released a blo….
0
7
0
@LiquidAI_
Liquid AI
1 day
The takeaway:. This is not a prototype. It’s a deployable, edge-ready solution for structured, secure, and fast function calling.
1
0
6
@LiquidAI_
Liquid AI
1 day
Real-world demos include:. - Creating reminders.- Setting projector displays.- Closing blinds to a set percentage. All performed with deterministic, low-latency function calls.
1
0
4
@LiquidAI_
Liquid AI
1 day
Compared to JSON:. `play_media(device="speaker", source="spotify", volume=1)`. This format uses 2.6x fewer tokens, generates faster, and is more reliable. It’s also more natural for LLMs trained on code.
1
0
8
@LiquidAI_
Liquid AI
1 day
Our stack:. LFM2-350M — a 350M parameter model optimized for <1GB RAM and <100ms latency. dotgrammar — a context-free grammar engine that ensures correct, parseable function outputs.
1
0
5
@LiquidAI_
Liquid AI
1 day
Why it matters:. Edge AI assistants in smart homes, vehicles, and mobile devices must act fast and correctly. Retry logic is not an option. Outputs must be right the first time.
1
0
6
@LiquidAI_
Liquid AI
1 day
We’ve partnered with @dottxtai and @remilouf to make function calling viable on the edge. JSON doesn’t cut it. We need structure, speed, and reliability.
Tweet media one
1
15
60
@LiquidAI_
Liquid AI
2 days
RT @TheTuringPost: 4 Outstanding families of models you must know. ▪️ Smol models by @huggingface.▪️ Qwen family + Qwen-Agent framework @Al….
0
21
0
@LiquidAI_
Liquid AI
9 days
RT @ramin_m_h: we will be having 3 grand hackathons soon at @LiquidAI_ offices in Boston, SF, and Tokyo. This is called “Bohemian Rhapsody,….
0
8
0
@LiquidAI_
Liquid AI
10 days
RT @duyluongdesign: I’ve been working with the @apolloaiapp team over the past few months to improve the app and launch a new local models….
0
5
0
@LiquidAI_
Liquid AI
11 days
RT @VentureBeat: Finally, a dev kit for designing on-device, mobile AI apps is here: Liquid AI's LEAP
Tweet card summary image
venturebeat.com
LEAP takes a local-first approach that allows small models to run directly on-device, no cloud infrastructure.
0
12
0
@LiquidAI_
Liquid AI
11 days
RT @localghost: Excited to share that Apollo has been acquired by Liquid AI. Expect more models, improvements, and powerful features you'….
0
8
0
@LiquidAI_
Liquid AI
11 days
RT @romankhrupa: Here’s a prototype I built in a few hours using the Leap SDK . It reads information from a website and uses a local LLM to….
0
2
0
@LiquidAI_
Liquid AI
11 days
RT @LiquidAI_: Today, we also release Apollo, an iOS native app for iPhone, Mac, and iPad, to test truly private, fast and AI in your hand.….
0
3
0
@LiquidAI_
Liquid AI
11 days
RT @xanamini: Anybody can run cloud LLMs — that's the past. Now with LEAP 🐸, you don’t need the cloud — just tap. No lag, no limits, no….
0
6
0
@LiquidAI_
Liquid AI
11 days
RT @jimmysmith1919: Extremely excited for our release of LEAP and Apollo. Our solutions for making it easy to develop apps powered by local….
0
2
0
@LiquidAI_
Liquid AI
11 days
RT @LiquidAI_: Today, we release LEAP, our new developer platform for building with on-device AI — and Apollo, a lightweight iOS applicatio….
0
46
0