I need a vacation @Meta
- Tweets 942
- Following 1,259
- Followers 10,789
- Likes 4,303
Pinned Tweet
time for some vacation stuff
Yesterday was my last day at @LumaLabsAI.
Over the last three years, I had the privilege of helping drive the company's transition from 3D AI to video generation and native multimodal foundation models.
I am grateful to have worked alongside an extraordinary group of researchers, and I look forward to seeing the next chapter of the company's story unfold.
9
1
182
32,381
Jiaming Song retweeted
5s of video, generated in 1.3s!
Nunchux brings @MiniMax’s MiniMax-H3 to @AMD MI355X with up to 26.7× faster inference than SGLang on 8 GPUs @AMDServer
And with streaming generation, you can change the prompt as the video plays, steering what happens next.
One stack, optimized across hardware.
Free access to MiniMax-H3 is coming soon! Join the waitlist now at nunchux.ai.
Our blog: nunchux.ai/blog/video-genera…
#AI #VideoGeneration #MiniMaxH3
13
27
121
100,604
Its kinda amazing how much @Muse can do in so little space (image was from yesterday and yes I saved $20 as a free user)
Introducing Muse, the personal agent that understands your goals and works 24/7 to get things done for you.
3
1
46
8,001
Jiaming Song retweeted
i collect vintage pens
when @alexandr_wang first sent me a TestFlight to this, i was in Amsterdam and just abt to go to bed
an hour later, i texted him with ‘holy shit’
it completely changed my travel plans the next day, and helped me find a rare pen i was in search of
SOTA
1/ today we're rolling out Muse, our new personal ai assistant. Muse is always-on, wicked fast, can use a browser, connect to your apps, and is designed to be secure.
try it now: muse.ai.
19
13
244
55,825
Jiaming Song retweeted
We're launching @muse, our new personal AI agent! Always on, fast, and built to be private, safe, and secure. Our biggest focus was the security architecture underneath. Each Muse runs in its own secure VM, a separate Sentinel checks every action, it never sees your passwords, and it asks before anything sensitive. More here: security.muse.ai
19
33
260
16,890
Uh oh
Many startup employees do not recognize the sheer litany of ways that a founder can screw you over without you even knowing. Founder trust is one of the most important things to look at when you’re joining a startup.
From cutting you out of M&A, screwing your retention pool, overdiluting your equity, firing you before your cliff, not having attractive options exercise plans, poor 409a price management, not telling you about QSBS / early exercise, blocking you from participating in secondary, obscuring company performance and many more.
Many many decisions that are made in rooms you are not in as an employee where the only thing that matters is: “does the founder have your back?” Great startups with untrustworthy founders lead to poor outcomes and often good startups with trustworthy founders lead to great outcomes.
Pick wisely.
2
32
16,712
Jiaming Song retweeted
I’ve had a lot of fun exploring Muse Spark 1.3 max!
Muse Spark 1.3 also brings exciting progress in visual coding, where advances in coding, agents, CUA, and multimodal converge to transform a spark of imagination into a vivid, compelling experience. Please give it a try!
1/ we just publicly released Muse Spark 1.3 max!
we see significantly stronger coding and agentic performance on muse spark 1.3 max, so would strongly recommend trying it out even if you've already tried muse spark 1.3 high or muse spark 1.3 xhigh.
2
4
65
16,420
Jiaming Song retweeted
1/ we just publicly released Muse Spark 1.3 max!
we see significantly stronger coding and agentic performance on muse spark 1.3 max, so would strongly recommend trying it out even if you've already tried muse spark 1.3 high or muse spark 1.3 xhigh.
141
167
2,017
332,343
Jiaming Song retweeted
Nunchux is live at nunchux.ai!
We’re building the frontier of multimodal generative AI inference: fast, affordable, high-quality serving for image, video, and world models.
Our Modelverse brings 30+ image, video, and avatar models behind one API. Use Nunchux-optimized models in Radical Speed and Radical Value tiers, alongside partner APIs. More models are on the way.
We’re opening access through the waitlist now. New accounts get $10 in credits.
Read our launch blog: nunchux.ai/blog/building-the…
39
48
195
223,292
Jiaming Song retweeted
Today we're releasing Muse Spark 1.3, our strongest model for agentic and coding tasks in the spark model line. It’s designed to support longer-horizon work, excel in agentic tasks, and follow complex instructions more reliably than previous models. Muse Spark 1.3 is available today in Muse Code and Meta Model API.
research.meta.ai/blog/introd…
12
50
448
38,933
Jiaming Song retweeted
this is a cool way to show progress
Wow!! we’re blown away by @Meta Muse Spark progress: 1.1 → 1.2 → 1.3.
Same building, same task of 3D simulation given only photos. The leap in geometry, structure, and visual fidelity is striking—even as the total cost stays at just $0.60.
Only 55 days from 1.1 to 1.3. Incredible pace, @AIatMeta! Congrats @alexandr_wang @shengjia_zhao @ren_hongyu etc! 🚀🚀
27
26
782
140,943
Jiaming Song retweeted
Muse Spark 1.3 is rolling out today with frontier performance almost too cheap to meter. This is the biggest jump we've made so far on coding and agentic work. Try it in Muse Code and our API.
Next up 🍉 and Muse Spark open weights releases coming soon.
1,446
1,617
19,961
3,763,661
Jiaming Song retweeted
Muse Voice Transcribe is MSL's first real-time audio perception model -- rolling out today. SOTA in streaming speech-to-text, it handles speaker diarization, and endpointing natively in a single model.
393
475
5,474
1,105,899
Jiaming Song retweeted
Muse Image is now available on Meta Model API and priced for production volumes at $0.01/image.
It’s an agentic image model that reasons before it renders. Each call searches the web to refine the output through iterative passes and evaluates the prompt for precise elements like charts and QR codes.
Text-to-image, single-image and multi-image editing, and multi-reference composition all live in one model, so there's no multi-step pipeline to stitch together.
Start building → bit.ly/4gsPeOV
38
72
626
137,269
Jiaming Song retweeted
Chelsea’s goal animation painful to watch, so here is how I would edit it.
24
24
1,130
111,776
Jiaming Song retweeted
many great ai companies about to be desperate because their growth is limited by compute
68
38
631
126,751
Jiaming Song retweeted
1/ muse spark 1.2 is a very strong multimodal model—it can do visual coding, robotics planning, and audio-visual understanding that all come together through agentic tools.
100
92
868
173,734
arxiv.org/abs/2608.17286
Abra provides the most controlled dense text-to-image estimate yet of the compute-optimal token/parameter ratio, and characterizes how deviations from that optimum, CFG, resolution, generation metrics, and learned representations behave.
@_selebou @skywalkeryxc @SwayStar123
2
12
169
24,454
