AI image generators

AI image generators turn a text prompt or a reference picture into a finished image. This grid compares popular tools side by side on what decides a choice: what you can do for free, what the paid plans really include, which model powers each one, whether it edits and keeps characters consistent, and what rights you get to use the results commercially. The real trade-offs are quality and control against price, and ease of use against how open and legally safe the output is.

Last updated

How to choose

Start with what you will make and how you will use it, then compare rights and cost before you compare image quality.

1. Decide what you need

  • Social posts and quick drafts: an assistant you already use or a free tier is often enough.
  • Brand work, product shots and series: you need Reference images, Character consistency and Image editing.
  • Posters, ads and slides: check Text in images, because many models still misspell words.
  • Print or large displays: compare Max resolution and Upscaling.

2. Rule out deal-breakers

  • Business use: check Commercial use and IP indemnification. Some free tiers do not allow commercial use.
  • Confidential or client work: check Public gallery by default, since some tools publish free-plan images.
  • Need to run it yourself or fine-tune it: only Open weights models qualify.

3. Compare the real cost

  • Plans count images in credits, GPU time or fast generations, so compare Plan allowance, not just the monthly price.
  • Developers should compare Pay per image API price, which depends on size and quality settings.

4. Judge quality fairly

  • Sample images on makers' sites are cherry-picked. Try the same three prompts, including one with text and one with people, in two or three tools.
  • The Independent Elo score comes from public voting and is one outside measure of quality.

5. Before you pay

  • Annual plans are cheaper per month but lock you in while models change every few months.
  • Check whether generated images carry Content credentials if platforms or clients require disclosure.

Answer

Browse the table. Filter to what matters.

Use the filters below, then sign in if you want to shortlist and decide with friends.

Sign in to vote

Sign in to vote, hide services and features, and reorder.

Verified: hover a cell and click its green dot for the source Estimate: no cited source Not stated: the company doesn't publish it

FeatureService ⠿ ⠿ ⠿ ⠿ ⠿ ⠿ ⠿ ⠿ ⠿ ⠿ ⠿ ⠿
Links
⠿ ⠿ ⠿ ⠿ ⠿ ⠿ ⠿
⠿
Free plan
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
Yes
Yes
⠿
Free plan limits
Image creation included, with separate usage limits; cap not published
Daily usage quota; downloads at 1K without a Google AI plan
Limited Imagine image generation; no video; exact cap not published
10 free Fast-mode creations, unlimited Standard mode; more with Microsoft Rewards
Free for everyday use; more generation needs a Meta One plan
Free daily generations across image, video and audio; refreshes daily
Limited monthly allowance for Standard and Premium AI tools; no Ultra AI
Doesn't apply
30 credits a week (Google/Apple sign-up); all images public
150 fast tokens per day; Auto model only; creations public
Limited credits; 3 uploads a day, 2 images per prompt; public, no commercial use
100 compute units per day
20 AI images per day (in-house model); personal use with attribution
5 images every 3 hours, 6 monthly credits, limited model access
Doesn't apply
25 free credits on signup
Doesn't apply
Free in Qwen Chat; API trial of 10 images per model for 90 days
Free open weights to self-host; needs several 80 GB GPUs
⠿
Cheapest paid plan
$8/month
$4.99/month
$8/month
$9.99/month
$7.99/month
$9.99/month
$18/month
$10/month
$20/month
$12/month
$12/month
$9/month
Not stated
$15/month
Doesn't apply
Doesn't apply
Doesn't apply
$6/month
Doesn't apply
⠿
Top individual plan
$500/month
$199.99/month
$200/month
$99.99/month
$19.99/month
$49.99/month
$25/month
$120/month
$60/month
$60/month
$160/month
$105/month
Not stated
$45/month
Doesn't apply
Doesn't apply
Doesn't apply
$43/month
Doesn't apply
⠿
Plan allowance
Go: 10x the Free tier's image creation; exact cap not published
2x higher usage than Free (Google AI Plus)
Base weekly usage pool shared across Grok and Cursor (Plus)
Higher usage than free for AI image creation in Copilot (Microsoft 365 Personal)
Meta One Core: more image and video generation than free; cap not published
2,000 generative credits per month (up to 400 images) plus unlimited standard Firefly image generations
Pro: 10x the Free plan's monthly AI allowance
3.3 hr Fast GPU time per month
2,400 priority credits per month (Plus), plus unlimited slow credits
8,500 fast tokens per month
1,000 credits per month (Basic)
5,000 units per month (about 200 Nano Banana 2.1 generations)
240,000 credits per year (Premium; credits valid 1 year)
1,500 monthly credits across all models (Pro, 2K); 6,000 on Pro Plus with 4K
Pay-as-you-go credits; no subscription
Pay-as-you-go credits at $0.01 each
Prepaid 30-day image packs from 700 images (Seedream 5.5); no subscription
20,000 credits per month shared across models (Token Plan Lite)
Doesn't apply
⠿
Pay per image API price
$0.02469/image
$0.034/image
$0.04/image
$0.019/image
$0.01/image
Not stated
Not stated
Doesn't apply
$0.03/image
Not stated
$0.035/image
$0.03/image
Not stated
Doesn't apply
$0.008/image
$0.08/image
$0.02/image
$0.03/image
$0.024/image
⠿
Current model
GPT Image 2.6 (Sunburst and Flare)
Nano Banana 2.1 (Gemini app uses Nano Banana 2; Pro also offered)
Grok Imagine Image 2.0
MAI-Image-2.6-Flash
Muse Image
Firefly Image Model 5
Canva AI Text to Image (runs on unnamed leading third-party models)
Midjourney V8.2
Ideogram 4.5
Lucid Origin
Recraft V4.1
Krea 3
Magnific One (plus 30+ partner models such as Nano Banana 2, Seedream 5 Pro)
Muse Image (default); also GPT Image 2.5, Nano Banana, Seedream
FLUX 3 Image (also FLUX 3 Image Klein, FLUX.2)
Stable Image Ultra (Stable Diffusion 3.5 Large)
Seedream 5.5
Qwen-Image 3.0 (and 3.0 Pro)
HunyuanImage 3.5 Preview (hy-image-3.5-preview); hy-image-3.0 API retired 2026-11-12
⠿
Text in images
Basic
Strong
Strong
Strong
Strong
Basic
Not stated
Basic
Strong
Strong
Strong
Not stated
Strong
Strong
Strong
Strong
Strong
Strong
Not stated
⠿
Image editing
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
⠿
Reference images
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
⠿
Character consistency
No
Yes
Yes
No
Yes
Yes
Not stated
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
No
Yes
Yes
Not stated
⠿
Max resolution
4K (3840 px max edge); above 2560x1440 experimental (API)
4K (API); 2K downloads in Gemini app with a Google AI plan
2K (API)
1.5K (MAI-Image-2.6 family)
1600 px (Meta Model API figure)
4 MP native; Upscale API for more
Up to 2K (Canva Craft model)
2048 x 2048
8192 x 8192 (8K, upscaled)
105MP (upscaled)
2K (Pro models)
22K (paid upscalers)
Up to 16x upscaling
2048 x 2048 (Pro); 4K (3840 x 2160) on Pro Plus
4K (about 16 megapixels)
4K (via Creative Upscaler)
Not stated
2K
Not stated
⠿
Upscaling
No
Yes
No
No
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
No
No
⠿
Commercial use
Yes on all plans
Yes on all plans
Yes on all plans
Limited
Not stated
Yes on all plans
Yes on all plans
Yes on paid plans
Yes on all plans
Yes on all plans
Yes on paid plans
Yes on paid plans
Yes on paid plans
Yes on all plans
Yes on paid plans
Yes on all plans
Yes on paid plans
Not stated
Limited
⠿
IP indemnification
Yes
Yes
No
No
No
Yes
Yes
No
No
No
No
Yes
Yes
No
No
No
No
Not stated
No
⠿
Open weights
No
No
No
No
No
No
No
No
Yes
No
No
No
No
No
Yes
Yes
No
No
Yes
⠿
API available
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
No
Yes
Yes
Yes
Yes
Yes
⠿
Video generation
No
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
No
Yes
No
No
Yes
Yes
⠿
Content credentials
Yes
Yes
Yes
Yes
Yes
Yes
Not stated
Yes
Not stated
Not stated
Not stated
Not stated
Not stated
Not stated
Yes
Yes
Yes
Not stated
Not stated
⠿
Web app
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
Yes
⠿
Mobile apps
iOS and Android
iOS and Android
iOS and Android
iOS and Android
iOS and Android
iOS and Android
iOS and Android
iOS and Android
iOS only
iOS and Android
iOS and Android
iOS only
iOS and Android
iOS only
None
None
None
iOS and Android
None
⠿
Public gallery by default
Private by default
Private by default
Not stated
Private by default
Private by default
Private by default
Private by default
Plan dependent
Plan dependent
Plan dependent
Plan dependent
Private by default
Private by default
Private by default
Not stated
Private by default
Private by default
Private by default
Private by default
⠿
Independent Elo score
1,261 points
1,176 points
1,167 points
1,124 points
1,126 points
Not stated
1,158 points
Not stated
1,017 points
877 points
1,027 points
Not stated
Not stated
1,126 points
1,115 points
Not stated
1,134 points
1,094 points
972 points

Why these features matter

  • Free plan

    Free access is the easiest way to test quality.

  • Free plan limits

    Free tiers often limit volume, resolution or commercial use.

  • Cheapest paid plan ($/month)

    The entry cost for regular use.

  • Top individual plan ($/month)

    Heavy generation and fast queues need the top tier.

  • Plan allowance

    Shows how many images you really get for the money.

  • Pay per image API price ($/image)

    The cost at scale for developers.

  • Current model

    Quality changes a lot between model versions.

  • Text in images

    Posters, logos and slides need legible text.

  • Image editing

    Edits save regenerating the whole image.

  • Reference images

    Needed for consistent brand style and characters.

  • Character consistency

    Needed for stories, comics and marketing series.

  • Max resolution

    Print and large displays need more pixels.

  • Upscaling

    Boosts resolution without regenerating.

  • Commercial use

    Business use needs clear rights.

  • IP indemnification

    Reduces legal risk for businesses.

  • Open weights

    Lets you run it locally and fine-tune it.

  • API available

    Needed to build your own app.

  • Video generation

    One subscription for images and video.

  • Content credentials

    Helps platforms and audiences identify AI images.

  • Web app

    Easiest access with no installation.

  • Mobile apps

    Matters if you create on your phone.

  • Public gallery by default

    Matters for confidential or client work.

  • Independent Elo score (points)

    An independent, vote-based quality score.

About the services

  • ChatGPT Images (GPT Image 2.5) · OpenAI

    OpenAI's image generation inside ChatGPT, with the GPT Image 2.6 models (Sunburst and Flare) also available through the API. Suits people who already use ChatGPT and want conversational editing; image caps on free and low plans are not published.

  • Gemini (Nano Banana) · Google

    Image generation and editing in the Gemini app and API with Google's Nano Banana models (2.1, 2, 2 Lite and Pro). Cheap entry through Google AI Plus, multiple reference images and character consistency; the app downloads 2K only with a paid Google AI plan.

  • Grok Imagine · xAI

    Grok Imagine Image 2.0 from xAI (now SpaceXSI), available in Grok through The X Subscription and by API at a flat per-image price. Edits with up to 5 reference images; consumer image limits are tied to a shared weekly usage pool.

  • Microsoft Image Creator (Bing / Designer) · Microsoft

    Microsoft's free Bing Image Creator, running MAI-Image-2.6-Flash, with extra generations through Microsoft Rewards; Designer and Copilot offer more image usage with Microsoft 365 plans. Best for casual or low-budget use; it is a consumer tool with limited commercial rights.

  • Meta AI (Muse Image) · Meta

    Meta AI's image generation and editing with Muse Image, free inside Meta AI and its apps. Meta One paid tiers ($7.99 and $19.99) are only in limited test markets, so US users get the free capped tier; no public API price.

  • Firefly · Adobe

    Adobe's image, video and audio studio with Firefly Image Model 5 (native 4 MP), Instruct Edit, upscaling and custom models for brand style. Suits designers and marketers who want Adobe's commercial-use positioning and Creative Cloud integration; credits are metered.

  • Canva AI image generation · Canva

    AI image generation (Dream Lab, with Canva's own Canva Craft model) built into the Canva design editor, so images drop straight into posts, slides and print layouts. Suits non-designers already on Canva; AI use is allowance-limited and grows with the plan.

  • Midjourney · Midjourney

    Subscription-only image generator known for stylised, artistic output, used through its web app. Better for concept art and mood boards than precise text or logos; no free plan, and companies over $1M revenue must buy the top tiers for commercial use.

  • Ideogram · Ideogram

    Image generator built around readable text, logos and posters, with a free weekly credit allowance and plans from $20 a month. Free images are public; private generation needs a paid plan.

  • Leonardo.Ai · Leonardo.Ai

    All-in-one creative platform with its own Lucid and Phoenix models plus partner models such as GPT Image, FLUX, Ideogram and Seedream, and video tools. Suits game, design and marketing teams that want many models in one subscription; allowances are in tokens, so cost per image varies by model.

  • Recraft · Recraft

    Design-focused generator with native vector (SVG) output, brand styles from reference images and mockups. Good for logos, icons and marketing graphics; the free plan has no commercial rights and its images are public.

  • Krea · Krea

    Creative suite combining its own Krea 3 image model with many partner image and video models, an upscaler, realtime generation and LoRA training. A free daily allowance and plans from $9 a month; credit-like compute units are shared across image and video.

  • Magnific (formerly Freepik AI Image Generator) · Magnific

    Freepik has been renamed Magnific: an AI suite (image, video, audio, upscaling) bundled with a large stock library and a commercial AI licence. Annual credit pools; paid tiers advertise unlimited generations on selected models only. Prices in this grid are not confirmed on the official page.

  • Playground · Playground

    AI design studio for logos, social posts and merchandise, running its own Muse Image model plus partner models such as GPT Image and Seedream. Free tier is capped at 10 images every 3 hours; paid plans add credits, upscaling and 2K-4K editing.

  • FLUX · Black Forest Labs

    Developer-first pay-per-image API and playground for the FLUX models (FLUX 3 Image, open-weights Klein, FLUX.2, editing tools). No subscription for consumers; best for builders who need control or self-hosting under a licence.

  • Stable Diffusion API · Stability AI

    Stable Diffusion 3.5 models with a credit-based developer API, editing and upscaling tools, and self-hosted community licensing. Aimed at developers and enterprises; no consumer subscription price could be confirmed.

  • Seedream · ByteDance

    ByteDance's image model (now Seedream 5.5) sold through the BytePlus API from about $0.02 per image, strong on infographics and multi-language text. Developer-oriented: pay per image or prepaid image packs, no personal subscription.

  • Qwen-Image · Alibaba Cloud

    Alibaba's Qwen-Image 3.0 and 3.0 Pro models via the Model Studio API at about $0.03-$0.04 per 1K image, with a free quota for new Singapore-region users. The 3.0 line shipped without open weights; consumer access is through Qwen Chat.

  • HunyuanImage · Tencent

    Tencent's HunyuanImage: the open-weights 80B HunyuanImage 3.0 and 3.0 Instruct, plus the newer 3.5 Preview on the Tencent Cloud API. Self-hosting needs several 80 GB GPUs; hosted access is aimed at developers rather than US consumers.

Suggest a change

You’ll sign in or create a free account to send it. We’ll keep what you typed.

Related comparisons