The each::labs blog. Models, guides and what we ship.
Model launches, tutorials, customer stories and product updates from the each::labs team.
Image Generation APILuma Uni-1: The Image Model That Thinks Before It Draws
Luma Uni-1 reads your brief like a model that understands images, then draws. What it does with nine references, directable edits and manga, and when to reach for Max.
Oct 22, 2026 · 8 min read
Video Generation APIPixVerse Swap: Replace Anyone in a Video, Step by Step
PixVerse Swap keeps the motion and changes who's in it. A step-by-step guide to clips that swap cleanly, replacement images that match, and running it on each::labs.
Oct 22, 2026 · 8 min read
Best OfBest AI Lip Sync Models in 2026: Pick by Your Footage
Lip sync isn't one job. Fixing real footage, animating a photo and turning a script into a presenter need different models. Here's which partner model to pick for each, all on each::labs.
Oct 22, 2026 · 8 min read
Image Generation APIKrea 3 Is Out of Beta: 8 Things to Know About Ultra
Krea 3 left beta on October 20 with Ultra mode in the API and a stack of leaderboard results. Eight things to know, from how Ultra picks its best attempt to when fast still wins.
Oct 22, 2026 · 7 min read
Video Generation APIKling 4.0 Is Out: 9 Things to Know Before the API
Kling 4.0 is live on Kling's own site, but the developer API is still marked soon. Nine things the launch actually confirmed, from depth-map references to the features still waiting.
Oct 20, 2026 · 8 min read
Text to Speech APIInworld TTS 1.5: 8 Things to Know Before You Build
Inworld TTS 1.5 is built for the half second that makes or breaks a voice conversation. Here are eight things to know about Max vs Mini, voices, formats and timing before you build.
Oct 20, 2026 · 9 min read
Image Generation APIFlux 3 Image: Edit One Thing, Keep the Rest
Black Forest Labs' Flux 3 Image places elements with boxes, edits one thing without touching the rest, reads real text and can search the web before it draws. Here's what changed since Flux 2.
Oct 20, 2026 · 9 min read
Use CasesFlux 3 Use Cases: 7 Jobs It Was Built For
Where Flux 3 actually earns its slot: banners that hold the grid, text in any language, product composites from ten references, step-by-step edits, grounded images and agent-written layouts.
Oct 20, 2026 · 9 min read
Video Generation APIOmniHuman 2.0: Five Minutes of Performance From One Photo
ByteDance's OmniHuman 2.0 animates a single photo to a full voice track of up to 300 seconds, with prompts for gesture and camera. Here's what changed from 1.5 and how to get good takes.
Oct 20, 2026 · 8 min read
Video Generation APIFlux 3 Video Keyframes: Pin Any Frame to Any Second
Flux 3 Video lets you pin up to ten images to exact seconds of a clip and generates one continuous shot through them. Here's how to plan the frames, write the prompt and finish the render.
Oct 20, 2026 · 8 min read
Video Generation APIGemini Omni 1.1 Flash Video Editing: Change One Thing
Google's Gemini Omni 1.1 Flash edits a clip from one plain sentence. The trick is saying less. Here's how to write edit prompts that change one thing and leave the rest of the shot alone.
Oct 20, 2026 · 7 min read
Music Generation APILyria 3.5: Full Songs From a Single Prompt
Thirty seconds of AI music is easy. A song with a verse, a chorus and a bridge that lands is not. Here's how Google's Lyria 3.5 holds one together, and how to prompt it on each::labs.
Oct 20, 2026 · 9 min read
Image Generation APIBest AI Image Editing Models in 2026: Who Edits Clean
Eight edit models, one question: how much of your image survives the edit? GPT Image 2.5, Nano Banana 2.1, Flux 3, Krea 3, Muse, Grok Imagine, Qwen and Ideogram 4.5 compared, plus a test you can run…
Oct 20, 2026 · 9 min read
Image Generation APIGPT Image 2.5 Use Cases: 7 Jobs for Flare and Sunburst
Seven jobs GPT Image 2.5 actually does well, from product listings to try-on and room edits, with a prompt for each and when to pick Flare or Sunburst.
Oct 16, 2026 · 8 min read
Video Generation APIKling 3 Motion Control: Copy Any Move, Step by Step
Kling 3 Motion Control copies a move from any reference clip onto your character. Here's the build order that keeps faces and limbs intact, from the first frame to the final upscale.
Oct 16, 2026 · 9 min read
PromptingKrea 3 Prompting Guide: Get Your Prompts Ready Now
Krea 3 is about editing with taste intact. Here's how to write prompts, reference sets and revision notes on Krea 2 today so launch day is a comparison, not a scramble.
Oct 16, 2026 · 8 min read
Image Generation APIMeta Muse Image: Ten References, One Clean Edit
Meta's first image model from its Superintelligence Labs builds one picture out of many references and leaves the rest of your image alone. It's live on each::labs.
Oct 16, 2026 · 8 min read
Image Generation APIKrea 3: Edits in Seconds That Keep Your Product Intact
Krea's new editing model changes what you ask for, keeps labels, faces and rooms where they were, and answers in seconds at native 2K. It's live on each::labs now.
Oct 16, 2026 · 7 min read
Video Generation APIVeo 3.1 First and Last Frame: Decide How a Shot Ends
Image to video tells the model where to start. Veo 3.1 first and last frame tells it where to land. Here's how to build both frames and the prompt in between.
Oct 16, 2026 · 8 min read
Music Generation APIMureka V9.5: AI Songs That Finally Hold Together
A great first verse is easy. A song that still makes sense at the bridge is not. Mureka V9.5 is built for the whole song, and it's live on each::labs.
Oct 15, 2026 · 7 min read
Video Generation APILTX 2.5: Multi-Shot Video, One Pass, Real Sound
LTX 2.5 writes the edit for you: connected shots with real cuts, synced sound in the same pass, and an audio-to-video route that lets your soundtrack direct the picture. Here's how to use all six…
Oct 15, 2026 · 9 min read
Video Generation APIHailuo AI: 9 Things to Know Before You Generate
Hailuo AI is an app, a model family and now H3. Here's which Hailuo to pick, how to prompt the motion it's built for, the bracket camera commands most people miss, and how to animate a photo.
Oct 15, 2026 · 9 min read
Video UpscalingTopaz Starlight Precise 2.6: AI Video That Looks Shot
Your AI clip has the shot. It just looks soft and a little plastic. Topaz Starlight Precise 2.6 is the diffusion upscaler built for exactly that last mile.
Oct 15, 2026 · 8 min read
Text to Speech APIXTTS v2: Clone a Voice From Six Seconds of Audio
Six seconds of someone talking, and XTTS v2 can say anything in their voice, in another language. Here's how Coqui's open voice cloning model works and how to get clean results.
Oct 15, 2026 · 6 min read