Skip to main content
MeiGen supports multiple AI models for image and video generation. Use the comparison table below to pick one, then scroll to its section for detailed specs.

​
Quick Comparison

ModelCreditsSpeedRef ImagesBest For
Z Image Turbo2Fastest (~5s)NoneQuick drafts, prototyping
GPT image1.52Fast (~20s)2Budget generation with references
Nanobanana 25Fast (~15s)5General use, wide ratio support
Seedream 5.0 Lite5Fast (~15s)5Photorealism, 3K output
Seedream 4.55Fast (~15s)5Product photos, marketing
Nanobanana Pro10Fast (~15s)5High quality, complex scenes
Niji 715Slow (~60s)1Anime, illustration, artistic
Midjourney V715Slow (~60s)1Photorealistic, universal aesthetics
Seedance 2.0Per-second (480p 13/sec, 720p 26/sec)Video, ~1-3minFirst/last frame + reference video4–15s adjustable duration, multiple ratios, audio
Veo 3.120 (fixed 8s)Slowest (2-6min)2 (first/last frame)8-second video with audio

​
Image Models

​
Nanobanana 2

ProviderGoogle (Gemini)
Credits5 per image
Prompt length5–4,000 characters
Reference imagesUp to 5
Resolutions2K, 4K
Aspect ratios1:1, 3:4, 4:5, 2:3, 9:16, 1:4, 1:8, 4:3, 5:4, 3:2, 16:9, 21:9, 4:1, 8:1
The most affordable Gemini model with the widest range of aspect ratios, including ultra-wide (8:1) and ultra-tall (1:8) options. Good for quick iterations and exploration.
Content moderation: Nanobanana 2 has stricter content filtering than other models. Prompts that mention celebrities, brand names, real people, or face-swap scenarios are more likely to be rejected. If your prompt is flagged, try rephrasing to avoid specific names or use a different model.

​
Nanobanana Pro

ProviderGoogle (Gemini)
Credits10 per image
Prompt length5–4,000 characters
Reference imagesUp to 5
Resolutions2K, 4K
Aspect ratios1:1, 3:4, 4:5, 2:3, 9:16, 4:3, 5:4, 3:2, 16:9, 21:9
The premium Gemini model with higher image quality. Better at complex scenes, fine details, and text rendering. Choose this when quality matters more than cost.

​
Seedream 5.0 Lite

ProviderByteDance
Credits5 per image
Prompt length5–2,000 characters
Reference imagesUp to 5
Resolutions2K, 3K
Aspect ratios1:1, 3:4, 4:3, 16:9, 9:16, 3:2, 2:3
ByteDance’s latest lightweight model. Offers good quality at low cost with optional 3K resolution output. Strong at photorealistic content and Chinese-style aesthetics.

​
Seedream 4.5

ProviderByteDance
Credits5 per image
Prompt length5–2,000 characters
Reference imagesUp to 5
Resolutions2K, 4K
Aspect ratios1:1, 4:3, 3:4, 16:9, 9:16, 3:2, 2:3, 21:9, 9:21
The previous generation Seedream model. Still produces solid results, particularly good for product photography and marketing visuals.

​
Midjourney Niji 7

ProviderMidjourney
Credits15 per image
Prompt length1–8,192 characters
Reference images1 (content or style reference)
Resolutions—
Aspect ratios1:1, 4:3, 3:2, 16:9, 3:4, 2:3, 9:16
Midjourney’s anime and illustration-focused model. Produces the highest quality artistic illustrations. Supports advanced parameters like stylize, chaos, weird, and style references. Unique features:
  • Content and style reference modes
  • Auto-translates non-English prompts to English
  • Advanced Niji 7 parameters (stylize, chaos, weird, raw, etc.)
  • Prompt Expand mode optimized for Midjourney syntax
Trade-offs:
  • Most expensive image model (15 credits)
  • Only 1 reference image supported
  • Slower generation (~60 seconds)
  • Returns 4 candidate images per generation (other models return 1) — all viewable in the image detail dialog

​
Midjourney V7

ProviderMidjourney
Credits15 per image
Prompt length1–8,192 characters
Reference images1 (content or style reference)
Resolutions—
Aspect ratios1:1, 4:3, 3:2, 16:9, 3:4, 2:3, 9:16
Midjourney’s general-purpose photorealistic model. Shares the same advanced parameters and reference image features as Niji 7, but targets universal aesthetics rather than anime style. Ideal for realistic photography, product shots, landscapes, portraits, and more. Same as Niji 7:
  • Content and style reference modes
  • Advanced parameters (stylize, chaos, weird, raw, etc.)
  • Auto-translates non-English prompts to English
  • Returns 4 candidate images per generation
  • Only 1 reference image supported
Different from Niji 7:
  • Default stylize is 0 (Niji 7 defaults to 100) — output stays closer to your prompt
  • Does not auto-append anime style keywords
  • Prompt enhance uses general polish mode (not Niji 7’s anime-specific expand)

​
Z Image Turbo

ProviderAlibaba (Qwen)
Credits2 per image
Prompt length5–4,000 characters
Reference imagesNot supported
ResolutionsFixed
Aspect ratios1:1, 2:3, 3:2, 3:4, 4:3, 7:9, 9:7, 9:16, 9:21, 16:9, 21:9
The fastest and cheapest model. Great for rapid prototyping and quick iterations when you need results in seconds.
No reference image support. Z Image Turbo generates images from text only — you cannot upload reference images with this model. If you need to use reference images, choose a different model.

​
GPT image1.5

ProviderRunningHub
Credits2 per image
Prompt length5–800 characters
Reference imagesUp to 2
Resolutions—
Aspect ratios1:1, 3:2, 2:3
An affordable model based on GPT image generation technology. Good balance of quality and cost with limited aspect ratio options. Limitations:
  • Shorter prompt length limit (800 characters)
  • Only 3 aspect ratios
  • Max 2 reference images

​
Video Models

​
Seedance 2.0

ProviderByteDance
CreditsPer-second pricing (see below)
Duration4-15 seconds (default 5)
Reference imagesFirst/last frame (image-to-video)
Reference videoYes (video continuation)
Resolutions480p, 720p (default 480p)
Aspect ratiosadaptive, 16:9, 4:3, 1:1, 3:4, 9:16, 21:9
AudioAuto-generated
ByteDance’s latest video generation model. Supports text-to-video, image-to-video (first/last frame), and video continuation. All outputs include AI-generated audio. The adaptive aspect ratio automatically matches the reference image/video dimensions.

​
Pricing

Per-second pricing, in two tiers: Without reference video (text-to-video / image-to-video):
ResolutionRate
480p13 credits/sec
720p26 credits/sec
With reference video (video continuation): lower rate, but with a minimum-billable-seconds floor.
ResolutionRate
480p8 credits/sec
720p16 credits/sec
For continuation, billable seconds = max(reference video duration + your duration, minimum billable). Lookup table:
Your duration (sec)Min billable (sec)
47
59
610
712
814
915
1017
1119
1220
1322
1424
1525

​
Pricing Examples

ScenarioCalculationTotal
480p × 5s text-to-video13 × 565
720p × 5s text-to-video26 × 5130
720p × 10s text-to-video26 × 10260
480p × 5s continuing a 3s video8 × max(3+5, 9) = 8 × 972
720p × 10s continuing a 5s video16 × max(5+10, 17) = 16 × 17272
If generation fails, the full credit amount is automatically refunded — including all per-second charges.

​
Veo 3.1

ProviderGoogle
Credits20 per video
Duration8 seconds (fixed)
Reference imagesUp to 2 (as first/last frame)
Resolutions720p
Aspect ratios16:9, 9:16
AudioYes (generated audio)
Google’s video generation model. Creates 8-second video clips with AI-generated audio. Supports speed control (Auto / Slow / Fast) and style presets (Auto / Realistic / Anime / Cinematic). Pick this for stable audio + style presets; for longer durations or more aspect ratios, use Seedance 2.0.

​
Known Limitations

This model is more sensitive to content involving real people, celebrities, brand logos, and face-swap requests. Prompts that work on other models may be rejected here. If you get a content violation error, rephrase your prompt to remove specific names or try a different model.
This model does not support reference images at all. The reference image upload section will be hidden when Z Image Turbo is selected. If you need to use reference images, switch to any other model.
Unlike other models that support up to 5 reference images, Midjourney models only accept 1 reference image. However, they offer more control over how the reference is used (content vs. style reference).
Unlike other models that return a single image, Midjourney models return 4 candidate images per generation. All candidates are saved and can be browsed in the image detail dialog. You are only charged once (15 credits) for all 4 images.
Only supports 3 aspect ratios (1:1, 3:2, 2:3) and has a 800-character prompt limit. Keep prompts concise with this model.
Both Seedream models cap prompts at 2,000 characters, shorter than the 4,000-character limit of Gemini models. Plan accordingly for very detailed prompts.
Unlike image models with flat per-image pricing, Seedance is billed per second: 13 credits/sec for 480p, 26 credits/sec for 720p. The sidebar shows the exact total before you submit.For video continuation: rates drop (480p 8/sec, 720p 16/sec), but a minimum-billable-seconds floor applies (so a short continuation isn’t unfairly cheapened by a long reference video). For example, continuing 5 seconds bills at least 9 seconds; continuing 15 seconds bills at least 25 seconds. See the full lookup table in the Seedance 2.0 section above.
Veo 3.1 always generates 8-second videos — duration is not adjustable. Only supports 16:9 / 9:16 aspect ratios and 720p resolution. For longer durations, more ratios, or 480p output, use Seedance 2.0.