SeeBloom — Turn one sentence into cinematic AI videos and images

Pick a creation type, then match the model — text/image to video, first→last frames, universal refs, edit, extend. Veo, Seedance, Kling, Wan and Nano Banana in one studio.

Create now · 30 free credits on sign-up

AI Video & Image Models

Pick a creation type, then match among 23 models from Google, OpenAI, ByteDance, Kling, Runway and more — text/image to video, first→last, universal refs, edit, extend.

AI video models

  • Veo 3.1 — Flagship AV · adherence / lip-sync / frames · 4/6/8s · 11cr/s (from 11 credits / second)
  • Veo 3.1 Fast — ~2× faster · first→last frames · 720p 3cr/s · 4/6/8s (from 3 credits / second)
  • Gemini Omni Fast — T2V/I2V 14 credits · edit 17 · 4–10s per request (14 credits / clip)
  • Grok Imagine — Per-sec: 480p 2cr · 720p 2cr · i2v +1 · edit +source (from 2 credits / second)
  • Runway Gen-4.5 — T2V / I2V · 2–10s · Standard 4cr/s (from 4 credits / second)
  • Runway Gen-4 Turbo — I2V only · 5s / 10s · Standard 2cr/s (from 2 credits / second)
  • Seedance 2.5 — Up to 30s · omni refs · first→last / edit / extend · 480p 4 / 720p 8 (from 4 credits / second)
  • Seedance 2 — Omni refs (image/video/audio) · 4–15s · 480p 2 → 4K 22 (from 2 credits / second)
  • Seedance 1.5 Pro — One reference image · 4–12s · 480p 1 / 720p 2 / 1080p 4 cr/s (from 1 credits / second)
  • Wan 3.0 — Omni refs · ≤10 img / ≤5 vid / ≤5 aud · first→last · native audio · 2–30s · 480p 2 / 720p 4 / 1080p 7 cr/s (from 2 credits / second)
  • Wan 3.0 Boundless — ≤10 img / ≤5 vid / ≤5 aud · first→last · Doc→Video · native audio · 2–30s · 480p 4 / 720p 8 / 1080p 14 cr/s (from 4 credits / second)
  • MiniMax H3 Max — Omni refs · ≤9 img / ≤3 vid / ≤3 aud · first→last · native audio · 5–15s · 480p / 768p 3cr/s (from 3 credits / second)
  • Kling 3.0 — T2V / I2V · fixed 5s / 10s only · 720p 3cr/s · 1080p 4 · 4K 12 (from 3 credits / second)

AI image models

  • Image-2 — 4K images with near-perfect text rendering (about 14 credits / image at default settings)
  • GPT Image 2.5 Sunburst — Highest-capability generation and precision editing (about 14 credits / image at default settings)
  • GPT Image 2.5 Flare — Faster high-quality everyday generate and edit (about 14 credits / image at default settings)
  • FLUX 2 Pro — Flagship photorealistic generation for production workflows (about 8 credits / image at default settings)
  • Nano Banana 2.1 — Next-gen Flash: sharper text, steadier characters (about 9 credits / image at default settings)
  • Nano Banana 2 — Pro quality at Flash speed (about 9 credits / image at default settings)
  • Nano Banana Pro — Google's flagship generation model (about 11 credits / image at default settings)
  • Seedream 5.0 Pro — Logically consistent images with intelligent visual reasoning (about 4 credits / image at default settings)
  • Kling Image — Kling text-to-image and image editing with flexible aspect ratios (about 3 credits / image at default settings)
  • Grok Imagine — xAI high-fidelity generate and edit, up to 2K (about 4 credits / image at default settings)

One account. Every top model.

No more separate subscriptions and top-ups on every platform. One credit balance covers Veo, Grok, Seedance, Kling, Runway, FLUX and more.

One-time purchase · No auto-renewal · Credits never expire

  • LITE: $2 for 40 credits
  • BASIC: $9 for 188 credits
  • PRO: $29 for 618 credits
  • MAX: $79 for 1888 credits
  • Works with every image & video model
  • Credits never expire
  • Failed generations refunded automatically

Frequently asked questions

What is SeeBloom?

SeeBloom is an AI video and image studio. Pick a creation type, then match the model — one account uses Veo 3.1, Seedance 2.5, Kling 3.0, Wan 3.0 Boundless, Runway Gen-4.5, Nano Banana 2.1 and GPT Image for text/image to video, first→last frames, universal refs, edit, extend, Doc→Video and AI images.

What video creation types does SeeBloom support?

Text/image to video, first→last frames, universal image/video/audio refs, video edit, video extend, and Doc→Video (PPT, PDF, Word, Excel or a public webpage). After you pick a type, the studio highlights models that support it.

What is Doc→Video?

Doc→Video on Wan 3.0 Boundless turns one PPT, PDF, Word or Excel file — or a public no-login webpage URL — into a short explainer clip. Document and webpage are mutually exclusive and cannot mix with first→last frames.

What is universal reference?

Universal reference lets you feed images, video and audio as constraints so the result tracks subject, pacing and sound. Caps differ by model — the studio shows limits for the model you pick.

Can I try it for free?

Yes. New accounts get 30 free credits — enough for a 10-second 480p Seedance 2.0 clip or several 2K images.

How does pricing work? Is there a subscription?

No subscription. Buy one-time credit packs ($2 = 40 credits, $9 = 188 credits, $29 = 618 credits, $79 = 1888 credits) with no auto-renewal. Credits never expire and work with every image and video model. The exact cost is shown before each generation.

Am I charged if a generation fails?

No. If a generation fails — including when it is blocked by a safety filter — the credits are refunded automatically and the reason appears in your notifications.

How long can my videos be?

It depends on the model: Veo 3.1 makes 4–8s clips, Kling 3.0 5 or 10s, Grok Imagine up to 15s and Seedance 2.5 up to 30s. For longer pieces, merge clips in your library or continue a shot with a model that supports video extend.

Can I write prompts in Chinese or English?

Yes. Write the prompt in whatever language you use. The interface follows the language you select.