fal

🚨 GPT Image 2 is live on fal, day 0! 🔤 Strong text rendering 🧭 Better layout + UI adherence 🛠️ Cleaner preserve-and-change edits 📷 Strong everyday photoreal output

Best open source world model for us to run as a public demo on fal.ai/wma? Comment it below ⏬ and let us show our unmatched scale!

fal-roster.vercel.app , webgl experience fully made using @fal the pipeline: GPT-Image-2 (or FLUX.2 Klein 9B for the live "create your own") 🦴 meshy v6 image-to-3d takes that → rigged GLB with a baked animation 🐾 GPT-Image-2/edit on the same character → companion creature image, color-locked to the character's palette → meshy v6 again (no rig) → static creature GLB 🪨 PATINA (fal-ai/patina/material/extract) takes the character image and outputs basecolor + normal + roughness + metalness for a per-character floor that matches their environment 🎬 Seedance 2.0 fast → 8s looping background video per character 📦 everything runs through gltf-transform (resize 1024 → webp q80 → draco) for ~95% size reduction → 400-800 KB GLBs 🖥️ the front-end is a single static index.html: Three.js scene, alpha-mapped marble disc with PlaneGeometry tessellation for real displacementMap depth, mirrored reflection plane, breathing animations on companions, palette-swap via CSS custom properties on character click 🧩 the whole recipe is published as a Claude Code skill fal-regenerate-3d so you can rebuild this yourself → github.com/fal-ai-communi… ⚠ metrics shown are decorative

. @burkaygur on creativity, craft, and what generative media actually changes for creators.

Try it now on fal: fal.ai/models/decart/…

Swap clothing, furniture or characters in real time.

Lucy 2.1 Virtual Try On is now live on fal! 🎥 Real-time video-to-video editing (live or recorded input) 👕 Swap outfits/characters live on camera 🔁 Preserves motion + scene continuity

fast inference world models

Try it on fal! fal.ai/models/fal-ai/…

🚨 Smart Resize is now live on fal. A multi-model pipeline, that takes any image + the dimensions you want and returns the same image at pixel-perfect size even when the aspect ratio changes. One endpoint. Any aspect. Same image, new dimensions.

Announcing World Model Accelerator, fal's new product: fal.ai/wma We've spent years building the inference layer for generative media. In 2023, we shipped 30 FPS, low-latency, action-controlled diffusion APIs, years before anyone else. Fast forward to today and we've been powering foundational model companies across image, video, real-time speech-to-speech and action-controlled world models. Now we're opening it up. World Model Accelerator is the system behind it all. Our in-house inference engine hits SOTA performance on Hopper and Blackwell for Diffusion Transformer workloads, both causal and bi-directional. Built on fal Serverless, it scales seamlessly from 1 GPU to 1,000+. This is the same platform that's been running real workloads, at scale, for years. A new WebRTC gateway optimizes latency between users and GPUs (routing requests to the closest region), built on the same infra that powers our real-time speech workloads. And direct access to the fal Marketplace puts your model in front of enterprises spending hundreds of millions on generative media, with a real co-selling motion behind it. This is the infrastructure layer for world models.

Try it on fal: fal.ai/models/nvidia/… fal.ai/models/nvidia/… fal.ai/models/nvidia/… fal.ai/models/nvidia/…

NVIDIA Nemotron 3 Nano Omni is now live on fal, available at launch. A single model for multimodal agents: 🔁 text, image, video, audio in one loop 🧠 1 context reasoning across complex workflows ⚡️ ~9× higher throughput with fewer inference hops Built for real-world agent systems.

Try it here now! Video Editing: fal.ai/models/alibaba… Reference to Video: fal.ai/models/alibaba… Text to Video: fal.ai/models/alibaba… Image to Video: fal.ai/models/alibaba…

Happy Horse 1.0 is live on fal, day 0 🐎 🎬 Best-in-class motion quality 🎧 Native 1080p with synced audio in one pass 🔗 Joint audio-video, not stitched 🔓 Fewer restrictions, broader commercial use ⚡ Built for production scale

We benched a few hundred prompts on Happy Horse 1.0. The short version: → ~20 words is the sweet spot → Subject, action, setting, one cinematography cue → Cut "stunning, beautiful, hyperrealistic" - they drag output to default → Camera moves earn extra length, put them last → Multi-beat? Use shot lists with timecodes, not prose Full guide on the fal learn 👇 fal.ai/learn/tools/pr…

Made this @fal-powered Roblox Studio plugin over the weekend, text/image to 3D using the new amazing Tripo P1 model! 🎮

@xai try it on fal 👇 fal.ai/models/xai/spe…

🚨 @xAI Grok Speech-to-Text is now on fal! 🎙️ 25+ languages with high accuracy 👥 Speaker diarization ⏱️ Word-level timestamps 🔊 Multichannel audio support