The each::labs blog. Models, guides and what we ship.
Model launches, tutorials, customer stories and product updates from the each::labs team.
Video Generation APIVeo 3.1 Lite: 9 Things to Know Before You Generate
Veo 3.1 Lite is the Veo you'll run most often. Nine things to know about its speed, durations, resolutions, sound and endpoints before you build it into your pipeline.
Nov 2, 2026 · 7 min read
PromptingImage to Video Prompts: Write the Motion, Not the Picture
Your image already answers who, what and where. A good image to video prompt only answers when and how. Here's how to write them, with examples for Kling 4.0, Seedance 2.5 and Wan 3.0.
Nov 2, 2026 · 8 min read
Text to Speech APIMiniMax Speech 2.8: Text to Speech That Breathes
MiniMax Speech 2.8 lets you write the breath, the pause and the laugh straight into the script. Here's what changed, HD vs Turbo, and how to run it on each::labs.
Nov 2, 2026 · 8 min read
Video Generation APILockstep: Cast Anyone, Keep Every Move
Lockstep, our own recast model, swaps anyone in a video from one photo each and keeps motion, lip sync and camera, for almost 75% less than other recast models.
Nov 2, 2026 · 9 min read
Video Generation APIKling 4.0 Reference to Video: Cast Once, Shoot Every Scene
Kling 4.0 reference to video takes up to 15 references in one request. Here's how to turn that budget into a series of shots that all look like one production, with prompts you can copy.
Oct 28, 2026 · 9 min read
Use CasesMiniMax Music 3.0: Songs That Keep Their Shape
MiniMax Music 3.0 is built for the part after the hook: full songs whose second chorus still sounds like the first. Here's how it works and how to brief it on each::labs.
Oct 28, 2026 · 7 min read
Best OfBest AI Upscalers in 2026: Faithful or Creative?
Some upscalers enlarge what's there. Others reinvent it. Here are the best AI upscalers for photos, AI images and video on each::labs, and how to pick the right one for the job.
Oct 28, 2026 · 8 min read
Image Generation APIBest AI Image Generation Models in 2026: Pick by Job
Leaderboards change weekly. Briefs don't. Here are the best AI image generation models of 2026, what each one is built for, and how to pick the right one for the job in front of you.
Oct 28, 2026 · 8 min read
Video Generation APIGenie 4 Is Coming: Google's World Model Learns to Talk
Google DeepMind's Genie 4 builds worlds you can walk through, steer with words and, for the first time, hear. Here's what's confirmed, what's missing, and how to get ready on each::labs.
Oct 28, 2026 · 7 min read
Audio Generation APIAdd Sound to AI Video: Music and SFX That Hit the Cut
Stock music never lands on your cuts. Here's how to add synced music and sound effects to AI video with Sonilo on each::labs, from a silent clip to a finished MP4.
Oct 28, 2026 · 7 min read
Video Generation APIFLUX 3 Max: Video Fast Enough to Iterate On
Black Forest Labs' FLUX 3 Max turns a ten-second clip around in roughly twelve seconds. Here's what that speed changes, how its text, image and keyframe models work, and how it compares with standard…
Oct 26, 2026 · 9 min read
Video Generation APIWan 3.0 vs Seedance 2.5: Which Video Model Wins Your Brief
Both make 30 seconds with sound in one pass. Wan 3.0 reads your deck; Seedance 2.5 wants your whole reference folder. Here's which brief belongs to which, plus prompts to run through both.
Oct 26, 2026 · 9 min read
Video Generation APIKling 4.0 Video Edit: Change the Shot, Keep the Take
You already have the take. Kling 4.0 Video Edit swaps the wardrobe, the background, the product or the whole style while your motion stays put. Here's every input, the workflow and five prompts to…
Oct 26, 2026 · 9 min read
PixVerseeach::labs × PixVerse: From Approved Master to Production
PixVerse V6 creates and approves the master video. each::labs checks its file and audio for the next stage.
Oct 23, 2026 · 7 min read
Image to 3DHunyuan3D 2.1: Turn One Photo Into a Textured 3D Model
Everyone can make the picture. Few teams can make the 3D asset. A step-by-step guide to turning one photo into a PBR textured model with Tencent's Hunyuan3D 2.1.
Oct 23, 2026 · 7 min read
Image Generation APIFIBO 2.0: Structured Prompts, Licensed Data, Same Shot Twice
Bria's FIBO 2.0 puts generation and editing in one model trained only on licensed data, and it reads scenes as structured JSON. Here's how brand and product teams can use it on each::labs to get the…
Oct 23, 2026 · 9 min read
Video Generation APISeedance 2.5 vs Kling 4.0: Two 30-Second Bets
Both models make thirty seconds in one take. Seedance 2.5 wants your whole folder; Kling 4.0 wants a director's plan. Here's how they differ and which brief belongs to which.
Oct 23, 2026 · 7 min read
Video Generation APIKling 4.0 vs Kling 3: What Changed and When to Switch
Kling 4.0 doubles the take and widens the control panel. Kling V3 is still the one you can call today. Here's what actually changed, and which shots belong to which model.
Oct 23, 2026 · 7 min read
PromptingHailuo AI Prompts: How to Direct Hailuo 2.3
Hailuo 2.3 got better at bodies, faces, physics and styles. Here's how to write prompts that actually ask for them, with MiniMax's bracketed camera commands and example prompts you can copy.
Oct 23, 2026 · 9 min read
Text to Speech APIGrok TTS: Text to Speech That Takes Stage Directions
Most text to speech throws away the margin notes. Grok TTS from xAI lets you write pauses, laughs, whispers and building intensity straight into the script, then voice it with Eve, Ara, Rex, Sal or…
Oct 23, 2026 · 9 min read
Video Generation APIMiniMax H3 Reference to Video: One Character, Every Shot
A reference pack, a prompt that gives every file a job, a shot list and a voice clip. Here's the step-by-step method for keeping one character identical across a multi-shot MiniMax H3 sequence on…
Oct 23, 2026 · 9 min read
Speech to Text APIDeepgram Nova-3: 9 Things to Know Before You Transcribe
Speech-to-text fails on the words that matter. Nine things to know about Deepgram Nova-3, from keyterm prompting and code-switching to redaction and summaries, before you send your first file.
Oct 22, 2026 · 8 min read
Image Generation APIGFPGAN: Restore Faces in Old Photos, Step by Step
Sharpening a blurry face gives you a sharper blur. Here's how GFPGAN rebuilds faces in old photos, which version to pick, and how to restore the whole picture on each::labs.
Oct 22, 2026 · 7 min read
Image Generation APIQwen Image 2.1: Transparent Layers and Ten-Reference Edits
The picture is the easy part. The parts are the job. Qwen Image 2.1 generates on transparency, combines up to ten references and edits exactly where you circle. Here's which version to use.
Oct 22, 2026 · 7 min read