Install this agent skill, inspect Flux fields, then run jobs through the RunAPI CLI.
Model Reference · CLI · SDK
Generate images with Flux Dev, Pro, and 2 Klein, or remix one image with Dev and Pro. This skill helps Claude Code, Codex, Gemini CLI, Cursor, and 50+ agents integrate Flux through RunAPI.
The canonical agent file is skills/flux/SKILL.md.
npx skills add runapi-ai/flux -gOr paste this prompt to your AI agent:
Install the flux skill for me:
1. Clone https://github.com/runapi-ai/flux
2. Copy the skills/flux/ directory into your
user-level skills directory (e.g. ~/.claude/skills/
for Claude Code, ~/.codex/skills/ for Codex).
3. Verify that SKILL.md is present.
4. Confirm the install path when done.
import { FluxClient } from '@runapi.ai/flux';
const client = new FluxClient();
const result = await client.textToImage.run({
model: 'flux-pro',
prompt: 'A cinematic product photo on warm paper',
aspect_ratio: '1:1',
});
const remix = await client.remixImage.run({
model: 'flux-pro',
prompt: 'Make this product shot feel like a warm editorial photo',
source_image_url: 'https://cdn.runapi.ai/public/samples/image.jpg',
aspect_ratio: '1:1',
});- Model page: https://runapi.ai/models/flux
- Product docs: https://runapi.ai/docs#flux
- SDK docs: https://runapi.ai/docs#sdk-flux
- SDK repository: https://github.com/runapi-ai/flux-sdk
- Pricing and rate limits: https://runapi.ai/models/flux/dev
- Provider comparison: https://runapi.ai/providers/black-forest-labs
- Browse all RunAPI models and skills: https://runapi.ai/models
- Integration work uses the target language SDK; one-off generation, manual smoke tests, debugging, or user-requested CLI runs use the RunAPI CLI skill: https://github.com/runapi-ai/cli-skill
- RunAPI-generated file URLs are temporary. Download and store generated images, videos, audio, or other files in your own durable storage within 7 days; do not treat returned URLs as long-term assets.
- Keep API keys in
RUNAPI_API_KEYor RunAPI CLI config; never commit secrets. - Prefer
create,get, andrunJSON passthrough patterns instead of inventing flags for every model parameter. - For pricing, rate-limit, and commercial-usage answers, link to the variant page rather than the repository README.
Licensed under the Apache License, Version 2.0.