AI Research engineer β Building apps like Scribe Instant Transcrips, VoiceView, ScreenRun β MTB π΅ ποΈ
Palo Alto and Chamonix
Joined April 2007
- Tweets7K
- Following2.5K
- Followers1.2K
- Likes17.5K
Laurent Denoue retweeted
T-minus 6.4 seconds.
That's how long our speaker diarization model, Who, took to label every speaker when we tested it on a 57-minute NASA podcast.
It ran at 530Γ real time, on a MacBook, with nothing leaving the device.
Laurent Denoue retweeted
It's insane that you can now run your own coding agent (Pi) w/ LFM3-1.7B *in the browser* at over 300 tok/s on an M4 Max, using custom WebGPU kernels & 4-bit GGUF weights.
No data sent to a server. All your chats stay private.
Try it yourself on Hugging Face π
Laurent Denoue retweeted
Introducing π»π΅πΌπΉ: our new generation of Liquid Foundation Models. With a new tri-hybrid architecture and new multimodal encoders, the LFM3 family is the ideal model suite for specialized intelligence.
> 6 model instances, across text, audio, vision, and omni modalities, generative and decision models.
> π»π΅πΌπΉ-π·.π½π±: text, 256k context window.
> π»π΅πΌπΉ-π
π»-πΈπ±: vision-language model with video-input support for spatial understanding
> π»π΅πΌπΉ-π°ππππ-πΈπ±: tool calls, voice chat, reasoning in 18 languages.
> π»π΅πΌπΉ-πΎπππ-πΈπ±: an omni model combines vision, audio, and language.
> ππ·-πΈπ± and ππ·-πΎπππ-πΈπ±: decision models with typed outputs and calibrated probabilities in a single pass. Omni adds audio input.
> fine-tune on 1 GPU.
> a new tri-hybrid architecture
> architecture optimized for 14 processor classes across GPUs, CPUs, and NPUs.
> π»π΅πΌπΉ-π
π»-πΈπ± is a beast. It performs competitively with Gemma 4 12B (6x larger), and Gemini 3.5 Flash Lite, and GPT-5.6 Luna, with 1.9B parameters.
> π»π΅πΌπΉ-π·.π½π± is #1 on the Pareto frontier of device intelligence πππππππ.ππππππ.ππ
Every modality for specialized intelligence. Pick the one that fits your use case, customize it, and run it on-device.
1/ π§΅
Incredible how optimizing image and video pre-processing to run RVM (background segmentation) LiteRT models can make your FSP jump from 6 (WASM) to 30 (LiteRT WebGPU) to 60 FPS (LiteRT WebGPU + optimized GPU video textures to avoid CPU roundtrips) ldenoue.github.io/rvm-webgpuβ¦
Nicegrab for macOS (clean screenshots and recordings with a simple key shortcut, composer over a background and smooth cursors) macmenubar.com/nicegrab/
100% browser based immersion of several video sources: recoloration based on ref image, relighting, segmentation, shadows, light sources and music beat detection driving the light intensity.
Now we just need to make it run at 60 FPS (possible on iOS) and of course RVM would produce much nicer mattes than the coarse masks given by MediaPipe (but RVM is slower!!)
Laurent Denoue retweeted
Today I'm releasing Paradee-8M-v1.1, a new version of my small open-weight text-to-speech model, distilled from Kokoro-82M. It's 8.5MB and it runs on any laptop without a GPU.
In v1.1, Paradee now runs 62x faster than real time on one CPU thread (compared to 18x from v1.0), and provides cleaner audio.
Laurent Denoue retweeted
Ray Dalio on the advice he gives his grandkids.
@RayDalio
Laurent Denoue retweeted
US: Tesla Roadster. EU: tethered bottle caps.
I made that joke too. But let me be serious for a moment.
Made in Europe.
Europe built it.
Others told the story.
The Web was born at CERN.
Linux in Helsinki.
Python in Amsterdam.
ARM, the chip in your phone and your Mac, in Cambridge.
DeepMind in London.
The first commercial microprocessor was designed by an Italian, Federico Faggin.
Every leading-edge chip on Earth is printed with ASML machines from the Netherlands.
Mistral trains frontier open models in Paris.
Hugging Face: French founders, although it is headquartered in New York.
llama.cpp, the engine of local AI, started in Bulgaria.
The first mRNA vaccine came from Mainz.
Europe doesnβt lack innovation.
It lacks a megaphone. π€·π»ββοΈ
Innovation without narrative becomes someone elseβs success.
This isnβt Europe vs. anyone.
I love innovation wherever it happens.
I just want Europeans to feel proud and optimistic about the future again.
Did I miss something relevant here?
app idea for macOS: voice control of my window, "show my xcode pdf project" because it takes so long to switch windows in this age of agent coding or "drag and drag huang.pdf onto chrome"
Laurent Denoue retweeted
Building apps with GPT-Live-1? Give them a voice that fits.
Hereβs how to shape a voice with a prompt or choose from new international preset voices:
Laurent Denoue retweeted
Apparently you need a cloud subscription to get accurate dictation π€‘
Bullshit.
$15/mo Β· Wispr Flow
$15/mo Β· Willow
$8.49/mo Β· Superwhisper
$8/mo Β· Aqua Voice
$0 Β· FluidVoice
Our Fluid-1 models run entirely on your device. Built for both speed and quality.
2,000+ tok/sec on an M5 Max
250+ tok/sec on an M1 Air 8GB
200+ tok/sec on iPhone.
Check us out below for Mac, iPhone and Windows!
Oh, if someone you know pays $15/month for dictation, send them this. And laugh π
Really good even with my french accent when speaking English. Congrats it's the fist of these kinds of apps I'll be keeping on my Mac (even the mini model, 256 MB)
Open weights, CC BY 4.0. Three builds:
- Core ML for iPhone and Mac
- ONNX,
- GGUF for Windows and Linux
Thanks @mudler_it for parakeet.cpp :)
huggingface.co/FluidVoice-Laβ¦
Full write-up with every number: fluidvoice.com/labs/fluid-spβ¦
Built on @NVIDIAAI Parakeet v3 and @moondreamai Parakeet Ultra. Thank you both :)
Promo code to download PDF Reflow for free for iOS macOS 4RNA9FNM7FNY appblit.com/pdfreflow
Laurent Denoue retweeted
How do always-on, proactive agents like @Bot work?
Watch my lecture at Stanford CS146S:
2:35 How we got here
6:13 What changed in the models
10:06 Inside an always-on agent
18:07 The harness
27:26 Context engineering
34:26 Where this is going