Krea 3 Ultra just landed at #4 on the Artificial Analysis image editing leaderboard.
1151 Elo. Nano Banana 2.1 is at 1152.
That's just one point.
Weird to see an indie lab in that lineup, huh? 😁
Style refs and LoRA training are next!
today, Krea 3 makes it to the top of image arenas.
- "Fast" pushes the pareto curve of speed and quality.
- "Ultra" ranks top-4, only surpassed by OpenAI models.
the only company valued under $1T in the top-10!
more below 👇
today, Krea 3 makes it to the top of image arenas.
- "Fast" pushes the pareto curve of speed and quality.
- "Ultra" ranks top-4, only surpassed by OpenAI models.
the only company valued under $1T in the top-10!
more below 👇
Making a good film trailer isn't easy.
Over the last few months, I built a skill for turning scripts, ideas, or loglines into trailers.
For this story, I started in July on the beach, with nothing to do but watch the sea.
This is GREEN WEEK.
Enjoy, workflow below 👇
There is only one model that produces high fidelity footage: Kling 4.0
You can have amazing texture and details.
I’ve had access to the full model for over a month and I tested it extensively.
Now, it’s available on Higgsfield, soon in 4K:
Condividi
Copiato
Higgsfield AI 🧩@higgsfield
Kling is back.
Kling 4.0 is now live on Higgsfield.
> Omni Reference: characters, objects and audio in one generation
> 30-second scenes in a single pass
> Start, middle and end frames to direct every beat
> Native 4K coming soon
Available now on Higgsfield.
The edit is code.
The score is measured beat by beat, and the cuts land on it.
Every line plays to its last breath, never clipped.
Independent AI reviewers then check the picture, sound, and cuts.
The final calls stay mine.
Kling is back.
Kling 4.0 is now live on Higgsfield.
> Omni Reference: characters, objects and audio in one generation
> 30-second scenes in a single pass
> Start, middle and end frames to direct every beat
> Native 4K coming soon
Available now on Higgsfield.
RT @maxescu: I gave Claude the Anthropic IPO files and asked it to make a film trailer about how we got here.
I present to you: "I Was There First"
Enjoy!
Step 1: Get a random product from Creative Market
Step 2: Tell Claude to use Higgsfield to create a promo video for the product
Step 3: Claude Code magic 🪄
Step 4: Profit!
Total time & cost: ~40mins and $20
You'd be crazy not to do this for your product!
I created a skill that creates high-quality brands.
Ideogram Vector generates the logo as an SVG.
It builds the contour and color studies from its paths, uses references in Higgsfield for stills and footage, then assembles the 30s film in a separate edit.
More details and workflow below 👇
I fed the product and lifestyle stills back into Higgsfield as image references.
The recorded Seedance 2.5 job requested 16s at 1080p.
The scene centered on pressing a body-wash pump.
I reviewed the output before editing it.
Krea 3 is one of the most fun models out there.
Smart, fast, and... available now.
Condividi
Copiato
Alex Patrascu@maxescu
Krea 3 Ultra just landed at #4 on the Artificial Analysis image editing leaderboard.
1151 Elo. Nano Banana 2.1 is at 1152.
That's just one point.
Weird to see an indie lab in that lineup, huh? 😁
Style refs and LoRA training are next!
For the simple body care test, I intercut 14s of SVG motion with 16s of Higgsfield footage, added an ElevenMusic v2.5 score, and exported a 30s 16:9 film.
Exact timing and vector overlays came from the edit.
HeyGen Video 1 now generates up to 30 seconds in a single take.
Any material, any style, one continuous shot at 2K.
$0.01/sec promo pricing through Oct 31.
Learn more:
Made with:
Claude Code Opus 5.5 Extra @claudeai
Higgsfield MCP @higgsfield_ai
Seedance 2.5, Nano Banana Pro
ElevenMusic v2.5 @elevenmusic_io
Skills:
music-video: my own pipeline for song analysis, lyric timing, cost ledger, and mux verification.
seedance-2-5: the house prompt format for every Seedance prompt.
built-in-browser: guidance for driving the in-app browser.
Local tools:
Demucs, faster-whisper, wav2vec2, MediaPipe Face Landmarker, librosa, PyTorch on CUDA and Hugging Face transformers, FFmpeg and ffprobe, yt-dlp, with Node.js, OpenCV, NumPy and Pillow, curl.
Python scripts I wrote for this run: prompt builder, assembler, voice map, word-lag sync, per-word residual check, and the B-roll mouth scan.
HeyGen Video 1 now generates up to 30 seconds in a single take.
Any material, any style, one continuous shot at 2K.
$0.01/sec promo pricing through Oct 31.
Learn more:
I still can't believe I can do this all by myself!
I vibecoded a new synth engine with a vintage twist in the browser.
All vibecoded with Argon High in Antigravity.
This is by far the most complex and best-sounding so far.
Play with it. Still WIP:
Hate all you want, but Jensen just put his money into Higgsfield.
He’s calling the founders visionaries and some of the brightest people in the world.
And TCV, the fund that backed Netflix and Spotify, is in too.
Condividi
Copiato
Higgsfield AI 🧩@higgsfield
Jensen Huang: “… the founders of Higgsfield are visionary in the work that they do … but these are some of the brightest people in the world.”
We're grateful to Jensen and the teams at NVIDIA and TCV for believing in us. Both companies have joined Higgsfield as investors in our Series B
extension.
Since our August Series B:
• We’ve crossed $1 billion in annualized revenue.
• Enterprise contracts are up 71%, growing 100%+ month-over-month
• We’ve grown from 30 million to over 40 million users globally.
TCV has backed Netflix, Facebook, ByteDance and Spotify, companies that changed how people create and consume media. NVIDIA has powered every generation of Higgsfield, from L40 and A40 GPUs through Hopper and Blackwell, with Vera Rubin next to meet growing demand.
Thank you to our creators and customers for building with us every day, and to everyone at Higgsfield whose work brought us here.
The full Bloomberg Announcement is linked below.
All I do these days:
1. Talk to Grok Bot to organize my life, both professionally and personally
2. Generate videos in Claude Code using Higgsfield MCP/API
3. Vibe code with Antigravity, using Argon High
4. Research and Computer Use with ChatGPT
I've made such a robust system that I feel like anything is possible.
Of everything on the list, the most mind-blowing thing lately has been Grok Bot.
I can't live without it now.
Jensen Huang: “… the founders of Higgsfield are visionary in the work that they do … but these are some of the brightest people in the world.”
We're grateful to Jensen and the teams at NVIDIA and TCV for believing in us. Both companies have joined Higgsfield as investors in our Series B
extension.
Since our August Series B:
• We’ve crossed $1 billion in annualized revenue.
• Enterprise contracts are up 71%, growing 100%+ month-over-month
• We’ve grown from 30 million to over 40 million users globally.
TCV has backed Netflix, Facebook, ByteDance and Spotify, companies that changed how people create and consume media. NVIDIA has powered every generation of Higgsfield, from L40 and A40 GPUs through Hopper and Blackwell, with Vera Rubin next to meet growing demand.
Thank you to our creators and customers for building with us every day, and to everyone at Higgsfield whose work brought us here.
The full Bloomberg Announcement is linked below.
Ingredients:
A link from Creative Market
Higgsfield MCP
Claude Code
ElevenMusic
Time:
~34min
Cost:
~$8-17
That's all you need to create a 30-sec showcase for your digital product.
In this case, a font:
RT @maxescu: X-Wave synth, completely vibe-coded in Antigravity with Gemini 4 Argon.
All done from scratch, just by talking in natural language.
Granular control, all in the browser.
Available here:
Have fun!
Claude Opus 5.5 + Higgsfield 👇
I had this idea earlier today while browsing Creative Market for fonts, and I thought:
"Why don't these fonts have a video showcase?"
The prompt was very simple:
"Make a 30-second showcase for this font: [LINK]. Figure it out and don't ask me anything. You have everything you need."
It took 55 minutes, and it cost just 481 credits in Higgsfield, which is ~$24
The music is also AI, made with ElevenMusic v2.5
RT @maxescu: My synth adventures continue...
This time, with a new synth from scratch: X-WAVE 80
Still WIP, but damn, it's fun making these!
And yes, it's still running 100% in the browser.
Gemini 4 Argon is a beast of a model:
I can’t wait to show you what I have planned with these
Condividi
Copiato
Alex Patrascu@maxescu
My synth adventures continue...
This time, with a new synth from scratch: X-WAVE 80
Still WIP, but damn, it's fun making these!
And yes, it's still running 100% in the browser.
Gemini 4 Argon is a beast of a model: