GIZMO / VIDEO

Every video model. One playground.

27 video models from Google, Kuaishou, ByteDance, MiniMax, Alibaba, xAI, Black Forest Labs, Lightricks, Luma, PixVerse, Shengshu. Text to video or image to video, with native sound on the models that make it, up to 30 seconds per clip. The same models work through the video API.

Open the video playground →Video API docs ↗

Video generation is part of Starter ($20/month) and above. Each clip is billed at the provider’s metered price plus your plan’s fee, after your weekly included usage.

Made with Veo 3.1“An elderly fisherman mends a net on a wooden dock at sunrise, gulls calling, he looks up and says "Good catch today." Cinematic, warm light”

The video models

Every clip below was made by the model itself. Open a model for its sample, price and every control it accepts.

GoogleVeo 3.1Cinematic shots with native sound, dialogue and realistic motion.4–8sAudioGoogleVeo 3.1 FastVeo 3.1 quality with audio at a lower price and faster turnaround.4–8sAudioGoogleVeo 3.1 LiteThe most affordable Veo, for drafts and high-volume clips.4–8sAudioKuaishouKling 3 ProFluid character motion and camera work with synced audio, up to 15 seconds.3–15sAudioKuaishouKling 3 StandardKling 3 motion quality for everyday clips.3–15sAudioKuaishouKling 3 4KNative 4K output from Kling 3 for large screens and fine detail.3–15sAudioKuaishouKling O3 ProKling’s reasoning video model: follows complex, multi-step direction.3–15sAudioKuaishouKling 2.6 ProProven, affordable Kling with audio in 5 or 10 second clips.5–10sAudioByteDanceSeedance 2.5Long multi-shot scenes up to 30 seconds with native audio.4–30sAudioByteDanceSeedance 2.0Consistent characters and smooth camera moves, up to 4K.4–15sAudioByteDanceSeedance 2.0 FastSeedance 2.0 at lower cost and faster speed.4–15sAudioByteDanceSeedance 2.0 MiniThe lightest Seedance, for quick social clips.4–15sAudioMiniMaxMiniMax H3 MaxMiniMax’s most capable video model: physics, faces and fine motion.1–15sMiniMaxMiniMax H3 Max TurboH3 Max quality at half the price and faster.1–15sMiniMaxMiniMax H3High-resolution MiniMax video up to 4K.5–15sMiniMaxHailuo 2.3 ProDynamic 1080p, six-second clips with strong prompt following.6sAlibabaWan 3.0Open-weight leader with audio, clips up to 30 seconds.2–30sAudioAlibabaWan 3.0 PrimeThe premium Wan 3.0 tier for the most detailed results.2–30sAudioGoogleGemini Omni Flash 1.1Fast, low-cost Gemini video with strong world knowledge.3–10sxAIGrok Imagine Video 1.5Quick, inexpensive video from xAI with lively motion.1–15sBlack Forest LabsFLUX 3 VideoPhotoreal FLUX look in motion, with audio, up to 20 seconds.5–20sAudioLightricksLTX 2.3 ProSharp 1080p to 4K video with audio, built for production.6–10sAudioLightricksLTX 2.3 FastLTX 2.3 speed tier, clips up to 20 seconds.6–20sAudioLumaLuma Ray 3.2Natural motion and lighting with keyframe control.5–10sPixVersePixVerse V6Stylized and social-ready clips at a very low price.1–15sAudioShengshuVidu Q3Anime and character animation with consistent subjects.1–16sAudioAlibabaHappy Horse 1.1Top-ranked text-to-video with vivid, coherent scenes.3–15s

Two ways to make video

Playground

Pick a model, write a prompt or drop in a start frame, set length, aspect ratio, resolution and sound, and generate. Clips keep rendering if you close the tab.

API

One request per clip at /api/v1/videos/generations. Poll for the MP4 and its receipt. GET /api/v1/videos/models lists every control and today’s price.

Claude & ChatGPT

Ask for it in a chat: connect Gizmo to Claude, ChatGPT or Codex over MCP. How to connect

Start creating →