Hedra
Log inSign Up
Open Hedra

AI image models

Hedra strives to support every state-of-the-art image model — and is constantly adding new ones as they become available.

Alibaba

Qwen Image 2.0Alibaba's Qwen-Image-2.0, tuned for speed. Native 2K output with professional in-image text rendering, for rapid iteration.

Black Forest Labs

Flux 1.1 ProPremium color depth and clarity when you want campaign-ready art that feels handcrafted.Flux 1.1 UltraThe big-canvas choice for ultra-high-res images and flagship visuals.Flux DevFast and light for quick concepts or high-volume social posts on a budget.Flux Kontext MaxHighest-fidelity reference-image support for complex, multi-element scenes and perfectly matched branded visuals.Flux Kontext ProReference-image support for character, brand, or style consistency.Flux.2 [flex]The tunable Flux.2 tier — trade denoising steps against speed per generation.Flux.2 [klein] 9BThe lean Flux.2 tier — quick, inexpensive stills for concepting and high-volume work.Flux.2 [max]The top Flux.2 tier, for realism and precision in final deliverables.Flux.2 [pro]The everyday Flux.2 tier — style transfer and sequential edits that hold together across passes.

ByteDance

Dreamina 3.1Polished, print-ready stills when the brief is a finished image rather than a sketch.Seedream 4.0Quick, reference-aware stills for professional work on a tight turnaround.Seedream 4.5Finer detail and steadier composition than 4.0, with support for several references at once.Seedream 5.0 LiteEdit or blend source images into a new 2K or 4K still, at the light tier's price.Seedream 5.0 ProPrecise edits to a source image with the top tier's multilingual in-image text, up to 2K.

Google

Imagen3Google's earlier photoreal generator, kept for parity.Imagen4Google's photoreal model—natural lighting, lifelike skin, and pro-grade sharpness in every shot.Nano BananaReference-guided stills that hold a character or product across a set.Nano Banana 2Multi-subject stills up to 4K — hand it several references and it keeps each one recognizable.Nano Banana ProThe reasoning-heavy tier — dense prompts, mixed references, and style transfer up to 4K.

HiDream

HiDream-O1-ImageHiDream.ai's open-weights HiDream-O1-Image (8B): one pixel-native model that generates, edits, and personalizes without a VAE or a separate text encoder.

Ideogram

Ideogram V2Best in class for poster-ready images and spot-on text rendering in social graphics.Ideogram V4Ideogram V4 renders poster-ready text and layout; the required quality parameter picks turbo, balanced or quality, which sets both the render effort and the price.

Microsoft

MAI-Image-2.5Microsoft AI's MAI-Image-2.5: photorealistic generation and editing with strong in-image typography and design-ready output.

NVIDIA

SanaLightning-fast and cheap for simple product shots or everyday content.

OpenAI

GPT Image 1.5Reads a long, specific brief closely — the choice when the prompt carries the detail.GPT Image 2OpenAI's balanced tier; moderate cost and fidelity, ideal for iterative refinement and everyday generation.

Recraft

Recraft V3Vector-clean graphics and crisp logos on demand.

Reve

Reve 2.1Generate images from a text prompt with strong prompt adherence, layout intelligence, and accurate text renderingReve 2.1 EditEdit one source image from a natural-language instruction, keeping the rest of the image intactReve 2.1 RemixCompose up to eight reference images into a new image from a text prompt

Topaz Labs

Topaz GigapixelUse the powerful and accurate Topaz image enhancer to upscale and enhance your images.Topaz WonderGenerative upscaling with realistic detail, precise text, and clean graphics — Topaz's highest-quality image upscaler.

xAI

Grok ImagineRework up to three source images from a plain-language instruction.Grok Imagine 2.0xAI's current Grok Imagine — the same irreverence at higher fidelity, from a prompt or from up to three source images.