Skip to content
Generate lifelike visuals with Seedream 5.5 ⚡

What’s new at MagicShot

A running log of new features, improvements and fixes, so you always know what’s changed.

Seedream 5.5 is now live

ByteDance's newest image model is here. Seedream 5.5 is built for long edit chains, so a photo stays clean through round after round of changes. It also renders small text more cleanly and delivers more realistic light and skin. Use it in Text to Image to create from a prompt, or in Image Editing to refine a photo you already have. Tap to try Seedream 5.5.

GPT Image 2.6 Sunburst is now live

OpenAI's high-detail image model is now on MagicShot. GPT Image 2.6 Sunburst is built for work where details matter, like brand and product imagery, dense layouts, and text-heavy designs. Its edge is consistency, so characters, products, and brand details stay intact edit after edit, and it generates about 50% faster than GPT Image 2.5. Use it in Image Editing for polished, precise results. Tap to try GPT Image 2.6 Sunburst.

GPT Image 2.6 Flare is now live

OpenAI's fast image model just got faster. GPT Image 2.6 Flare generates and edits images about 50% quicker than the previous generation, and it is built for rapid iteration, so you can try an edit, tweak it, and try again without the wait. Upload a photo in Image Editing, describe the change, and go. Great for quick drafts, social content, and high-volume work. Tap to try GPT Image 2.6 Flare.

Lyria 3.5 is now live

Google's newest music model is here. Lyria 3.5 turns a prompt into full songs with distinct verses, choruses, and bridges, with more expressive vocals, clearer pronunciation, and lyrics that follow your prompt more closely. Melodies are richer and sound more natural, and the model is built for more control over tempo and duration. Describe the song you want in Music Generator and let it sing. Tap to create with Lyria 3.5.

Gemini Omni 1.2 Flash is now live

Google's fast video model just got an upgrade. Gemini Omni 1.2 Flash now creates clips up to 15 seconds long, up from 10 in the previous version, with native synchronized audio. It follows your references more closely and delivers more natural speech with lip sync, including dialogue and multiple speakers. Type a scene in Text to Video and watch it come to life with sound. Tap to try Gemini Omni 1.2 Flash.

MagicShot MCP is live

Connect MagicShot to Claude, Claude Code, Cursor, VS Code, Codex or Windsurf over MCP and generate images and videos right from your chat or editor. Just ask in plain words. It runs on the credits you already have, and in Claude there's no API key to paste.
Connect MagicShot →

Nano Banana 2.1

Google's Nano Banana 2.1 just landed on MagicShot. It brings better visual design, cleaner mask-based edits that change only the area you mark, and steadier subject consistency across multiple edits, with output up to 4K. It's fast, and it costs less than Nano Banana 2, so iterating on an idea is easy on your credits.
Try Nano Banana 2.1 →

Ideogram 4.5

Ideogram 4.5 is now on MagicShot, and it's built for precise edits. Change just the part you ask for, like a misspelled word, a color, or a pose, while the rest of the image stays untouched, even after round after round of revisions. Great for product shots, posters, and campaign variations without starting over.
Try Ideogram 4.5 →

Flux 3 Image

Black Forest Labs' Flux 3 Image is now live on MagicShot, and it's built for control. Lay out where every element goes, edit one part of a finished image while everything else stays exactly as it was, or combine up to 10 reference images into one well-composed shot. Renders in native 4K, so text and fine detail stay sharp.
Try Flux 3 Image →

GPT Image 2.5 Flare

OpenAI's fast new image model just landed on MagicShot. GPT Image 2.5 Flare is built for speed, delivering higher-quality images than the last generation at up to 50% lower latency, so it's perfect for social content, quick concepts, and high-volume work. You get sharper detail, more natural lighting, precise edits that change only what you ask, and native transparent backgrounds. Its edge is turnaround: create, tweak, and repeat without the wait.
Try GPT Image 2.5 Flare →

GPT Image 2.5 Sunburst

GPT Image 2.5 Sunburst just landed on MagicShot. It's OpenAI's precision model, trading a little extra generation time for tighter control over every edit. Preserve a subject across restyles, revise one element without touching the rest, and stack multiple edits without the image degrading. Built for production-ready campaign and product imagery where quality can't slip.
Try GPT Image 2.5 Sunburst →

New kisses just unlocked

New styles just dropped in Kissing Booth. Upload your photo, pick a preset, and generate. From tender to playful, there are more ways than ever to bring a kiss to life on video. No prompt writing, just choose a look and go.
Try AI Kiss →

MiniMax H3

MiniMax H3 is an omni-modal video model built by MiniMax that reads text, images, video, and audio together and turns them into finished video clips. On MagicShot, it works as a text-to-video and image-to-video engine that renders motion, on-screen text, and native stereo sound in a single generation.
Try MiniMax H3 →

Flux 3.0

Flux 3 is Black Forest Labs' first video model, a multimodal system that generates clips up to 20 seconds long with native audio in a single pass. Inside MagicShot, it turns a text prompt, a still image, or keyframes into finished video with sound.
Try Flux 3 →

Seedream 5.0 Pro

A flagship image model that separates one render into 10+ editable PNG layers, reads spatial annotations, and renders in-image text across roughly 14 languages. Here's what it does and why it matters.
Try Seedream 5.0 Pro →

Seedance 2.0 is here

Our most cinematic AI video model yet. Generate longer, sharper clips from a single prompt.
Try Seedance 2.0 →

Ready to make your first one?

Make images, videos and voiceovers in one place.