UnitPic

AI Image Model Hub

One account for every leading AI image model. Grouped by provider, so you can find the best AI image model for your use case.

Every AI image model in this hub runs on the same credit balance — pick any AI image model, or run several AI image models on one prompt and compare the results side by side. Use the filters below to narrow the AI image model catalog by vendor, speed, price, and strength.

Not sure which AI image model to start with? Open any AI image model profile for its credits, latency and strengths, or head-to-head pages to see two AI image models judged on the same prompt. Every AI image model page links straight into Studio so you can try it on your own prompt.

Black Forest Labs

7 models
FLUX.1 Schnell1 credits
FLUX.1 Schnell

Ultra-fast image model, roughly 2 seconds per image, great for rapid iteration

FastGeneral
~2s★★★★☆
Use model
FLUX.2 Klein 4B2 credits
FLUX.2 Klein 4B

Lightweight FLUX.2 variant with balanced speed and quality

BalancedHigh quality
~3s★★★★☆
Use model
FLUX.2 Klein 9B2 credits
FLUX.2 Klein 9B

Ultra-fast distilled model with enhanced quality, ideal for live preview

High qualityReal-time
~4s★★★★☆
Use model
FLUX.2 Dev2 credits
FLUX.2 Dev

FLUX.2 flagship model with outstanding realism and detail

FlagshipMulti-reference
~15s★★★★★
Use model
FLUX.2 [pro]8 credits
FLUX.2 [pro]

BFL's recommended production-grade default model, supports 8 reference images

ProductionMulti-reference
~10s★★★★★
Use model
FLUX.2 [max]18 credits
FLUX.2 [max]

BFL's highest-quality model, supports Grounding Search

Top qualityGrounding
~15s★★★★★
Use model
FLUX.2 [flex]13 credits
FLUX.2 [flex]

Tunable steps/guidance, ideal for layout-heavy production workflows

FlexibleTunable
~10s★★★★★
Use model

Stability AI

1 models
S2 credits
Stable Diffusion XL Base 1.0

Classic open-source SDXL base model

ClassicOpen source
~8s★★★☆☆
Use model

ByteDance

5 models
SDXL Lightning2 credits
SDXL Lightning

Lightning SDXL that generates high-quality images in just a few steps

LightningOpen source
~3s★★★★☆
Use model
Seedream 5 Pro12 credits
Seedream 5 Pro

ByteDance's high-quality generation and editing model with 10 reference images

High qualityMulti-reference
~8s★★★★☆
Use model
Seedream 4.510 credits
Seedream 4.5

Multi-reference, batch and sequence generation

BatchMulti-reference
~7s★★★★☆
Use model
Seedream 5 Lite9 credits
Seedream 5 Lite

Lighter and faster, supports 14 reference images and batch generation

LightweightBatch
~5s★★★★☆
Use model
Seedream 4.08 credits
Seedream 4.0

Unified text-to-image and editing architecture, up to 4K resolution

Image editing4K
~6s★★★★☆
Use model

xAI

1 models
Grok Imagine Image 2.010 credits
Grok Imagine Image 2.0

xAI Grok image generation with a distinctive creative style and text rendering

CreativeText rendering
~5s★★★★☆
Use model

Alibaba

2 models
Qwen Image 3.0 Pro10 credits
Qwen Image 3.0 Pro

Alibaba Qwen flagship text-to-image model with strong prompt understanding and negative-prompt support

High qualityMultilingual
~5s★★★★☆
Use model
Wan 2.6 Image8 credits
Wan 2.6 Image

Tongyi Wanxiang 2.6 text-to-image model with rich detail and negative-prompt support

DetailValue
~5s★★★★☆
Use model

Google

5 models
Nano Banana Pro36 credits
Nano Banana Pro

Google's higher-quality version with better detail and prompt adherence

High qualityMulti-aspect
~5s★★★★☆
Use model
Nano Banana 218 credits
Nano Banana 2

Second-generation model with better quality and speed, supports search

Next-genSearch
~4s★★★★☆
Use model
Imagen 410 credits
Imagen 4

Google's latest photorealistic model

PhotorealRealism
~6s★★★★★
Use model
Nano Banana9 credits
Nano Banana

Google's fast text-to-image model, high quality with multiple aspect ratios

FastGreat value
~3s★★★★☆
Use model
Nano Banana 2 Lite9 credits
Nano Banana 2 Lite

The fastest Gemini image model, ideal for rapid iteration

Blazing fastIteration
~2s★★★★☆
Use model

OpenAI

2 models
GPT Image 210 credits
GPT Image 2

OpenAI's next-generation image model with editing support, no transparent backgrounds

Image editingMulti-quality
~8s★★★★★
Use model
GPT Image 1.510 credits
GPT Image 1.5

OpenAI image generation and editing with vivid/natural style support

Image editingStyle switch
~8s★★★★☆
Use model

Krea

2 models
K15 credits
Krea 2 Large

Large model, photorealistic, supports style references and mood boards

PhotorealMood board
~8s★★★★☆
Use model
K8 credits
Krea 2 Medium

Smaller and faster, excels at illustration, anime and painterly styles

IllustrationValue
~5s★★★★☆
Use model

Leonardo

2 models
L6 credits
Lucid Origin

Strongest prompt responsiveness with accurate text rendering

Text renderingFull HD
~8s★★★★☆
Use model
L6 credits
Phoenix 1.0

Excellent prompt adherence, ideal for precise control over the scene

Prompt adherenceCreative
~10s★★★★☆
Use model

Lykon

1 models
L2 credits
DreamShaper 8 LCM

Fine-tuned model balancing realism with broad stylistic range

RealismVersatile styles
~6s★★★☆☆
Use model

Pruna

1 models
P1 credits
P-Image

Ultra-fast text-to-image with automatic prompt enhancement and LoRA support

Ultra-fastAuto-enhance
~2s★★★☆☆
Use model