All tools
Catalog · 16 services
AI image generators
Widely-used image-generation services in 2026, with one-line honest assessments. Each entry links out to the service. None of these run locally in this app — for templates with text overlays, use our Visual Builder instead.
The indie-founder stack — if you only pick 3
- Midjourney v7 — for any hero / brand / mood imagery. Pay for it. It compounds.
- Ideogram 2.0 — only when you need text inside the image (posters, ad layouts).
- ChatGPT Plus (GPT Image 1) — already in your pocket if you have ChatGPT. Use for quick iteration in conversation.
What to NOT use AI imagery for
- Your actual product UI screenshots — show the real thing.
- Talking-head shots of you or your team — uncanny in 2026, signals low effort.
- Anything that needs precise text — except via Ideogram or our Visual Builder.
- App Store screenshots — use a Figma template instead (see design-automation).
Flagship general-purpose
The five you'll actually pick from for 90% of jobs. If you only learn one, learn Midjourney.
- Still the best for aesthetic, art-directed output. Web app is finally good (Discord-only era is over). No official API — work in browser, export.
Best for: Conversational refinement, multi-turn edits
SubscriptionConversationalStrong prompt adherence and great at iterating in conversation. Stylistic range is narrower than Midjourney. Bundled with ChatGPT Plus — no extra fee.- The free tier is generous and the text-rendering is genuinely best-in-class. Style range is corporate-clean — fine for marketing, lacks Midjourney's character.
Best for: Anyone wiring image gen into their own product
APIOpen-sourceBest open-weights family. Available via fal.ai and Replicate APIs — most indie SaaS that ships image gen uses Flux behind the scenes. Photoreal quality rivals Midjourney.- The only one that does proper vector + brand-style consistency across a set. If you need 10 illustrations that look like they came from the same designer, this is it.
Specialty
Each one is best-in-class for a specific job. Don't use them as your daily driver.
- When you need a poster, magazine cover, ad layout with real text rendered into the image. Most others mangle text; Ideogram doesn't.
Best for: Commercial-safe imagery, Photoshop generative fill
SubscriptionCommercial-safeTrained on Adobe Stock + licensed data — the only major model with a clean copyright story. Quality is mid-pack but the legal certainty matters if you're scaling.- Realtime generation while you sketch or adjust prompts. Best for the ideation phase — see 50 variations in a minute, then refine elsewhere.
- Big free daily allowance, broad model library, character-consistency tools that other platforms charge extra for. Quality varies by model — pick carefully.
- Not generation from scratch — takes a 512px image and turns it into a 4K masterpiece, or relights an existing photo. Expensive but uniquely capable.
Best for: Total control, self-hosting, fine-tuning on your style
Open-sourceSelf-hostRun locally via ComfyUI or Automatic1111. Steep learning curve but unmatched control — LoRA training on your brand assets is a real workflow once you commit.
Worth knowing about
Niche, newer, or accessible-but-not-our-default. Useful in specific cases.
- Multiple models (Mystic, Flux, Imagen) behind one UI with a generous free tier. Lower ceiling than a dedicated Flux or Midjourney workflow, but easy.
- Search a database of others' prompts + outputs. Use it to find the right prompt patterns before generating elsewhere. Underrated for prompt engineering.
- Newer entrant with very strong photorealism. Free tier exists. Watch this one — the trajectory is steep.
- Image+motion crossover. Useful if your next stop is animating the image (with Veo / Sora / Runway). Niche but growing.
- Not a model — a host. Flux, SDXL, Stable Diffusion variants behind a unified API. The most common starting point for developers shipping image gen in their own product.
- Same idea as fal.ai but with a long tail of community models. Slower-feeling but cheaper for some workloads.
Need to brief these tools systematically?
The social-image-spec skill writes a spec a developer can implement in Vercel OG. The figma-template-brief skill writes a spec for a Figma designer. Use these to drive image production at any scale.