The image model landscape — six families and what each is for — step 1 of 9

Six families, six jobs

You do not need accounts, APIs, or a design job to read this chapter. Each name below is a picture-making tool: you type a description, it returns an image. Midjourney is the one a lot of people know — a browser/Discord app where you describe a picture and it draws one. You do not need Midjourney (or any of these) installed to finish the drills here. Prices are a map, not a bill — skip the dollar rows if you just want which tool is for which job.

In 2024 there was effectively one mainstream image model people knew about (DALL-E 3) and a few power-user choices (Midjourney, Stable Diffusion forks). In 2026 the landscape has forked into six distinct families, each optimized for a different job. Picking the wrong family is the single biggest source of bad output in image generation today — bigger than prompt skill, bigger than budget.

The six families:

1. Flux (Black Forest Labs) — open weights, leads on photorealism and technical accuracy. The 2026 lineup is FLUX.2: flux-2-pro (from $0.03/MP) and klein (from $0.014). Previous-gen flux-pro-1.1 is still live at $0.04. API access via Black Forest Labs, Replicate, and Fal.ai.
2. nano-banana (Google Gemini Flash Image) — the rebranded marketing name for Gemini 3.1 Flash Image (Nano Banana 2, gemini-3.1-flash-image), Gemini 3.1 Flash Lite Image (gemini-3.1-flash-lite-image), and Gemini 3 Pro Image (Nano Banana Pro, gemini-3-pro-image). The default call is generate_content with an IMAGE modality, one image per call. Speed-optimized, cheap per image, strong multi-turn editing. Gemini 2.5 Flash Image at $0.039 retires October 2026; the old preview ID is already gone.
3. Midjourney — closed-source, no public API, Discord and web app only. Wins on subjective aesthetic taste; V8.x narrowed the old text-in-image gap but still trails the specialist for letter-perfect copy, and it loses outright on character consistency and any workflow that needs more than 200 images a month.
4. Ideogram — text-in-image specialist; 4.0 shipped June 2026. Claims 90-95% text rendering accuracy. Three tiers (as of September 2026): Turbo $0.03, Default $0.06, Quality $0.10. Endpoint: POST /v1/ideogram-v4/generate.
5. OpenAI gpt-image family — gpt-image-2 is the current Image API default. Square output only: quality=low $0.006 / medium $0.053 / high $0.211. Strong instruction-following (you can say "put the logo top-left at 20% size" and it usually listens). DALL-E 3 shut down May 2026. gpt-image-1 shuts October 2026; mini and 1.5 shut December 2026. This family has the fastest release cadence in the table — re-check the OpenAI Images API page for the current model name before you ship.
6. Recraft (V4.1) — vector and brand-mark specialist. Model ID recraftv4_1: $0.035 raster / $0.08 vector. If the brief needs an SVG, this is the family. Imagen 4 on the Gemini API shut down August 2026.

The two axes that actually matter

When you're picking a model, two axes determine 80% of the decision:

• Taste vs. control. Midjourney is taste-heavy: you describe a vibe and it gives you a magazine shot. Flux and gpt-image-2 are control-heavy: you describe a specific composition and they execute on it. If you have art direction, you want control. If you don't, you want taste.
• Per-image cost vs. quality ceiling. nano banana 2 at $0.067/image (1K) is the batch workhorse — $0.045 at 0.5K, $0.101 at 2K, $0.151 at 4K. Lite is $0.0336 at 1K. flux-2-pro from $0.03/MP is the photoreal workhorse. Midjourney Mega at $120/month is unlimited-Fast-GPU for one creator. gemini-3-pro-image at $0.134 (1K/2K) / $0.24 (4K) is the Google premium ceiling.

Everything else (character consistency, text-in-image, vector output) collapses into a specialty model: if you need it, use the specialist.

What Claude can and can't do

One trap before we go further. Anthropic's Claude (the model running this curriculum) reads images — you can pass {"type": "image", ...} blocks into a messages.create and Claude will analyze them. Claude does not generate images. If you wire up an image-generation feature, the call goes to Google, OpenAI, Black Forest Labs, Ideogram, Recraft, fal.ai, Replicate, or directly to a self-hosted Flux endpoint. Not Anthropic.

This trips up about half the people who try to ship an image feature in their first Claude app. Don't be one of them.

What this lesson does

Walks the six families with their actual 2026 names, prices, and strengths. Drills you on which one to pick for five different jobs. Ends with a pick_image_model(task) function you can call from any harness to route an image-gen request to the right backend. Pure decision logic — no API calls, runs in stdlib.

The downstream lessons assume you can tell, given a task description, which family you'd send it to. By the end of this lesson, you can.