fal-ai-media

Generates images, videos, and audio through fal.ai MCP models.

Updated Mar 25, 2026
One-click install
npx skills add https://github.com/Femad-6/my-skills --skill fal-ai-media-femad-6
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: fal-ai-media
Source: https://github.com/Femad-6/my-skills/tree/main/.github/skills/fal-ai-media
Command: npx skills add https://github.com/Femad-6/my-skills --skill fal-ai-media-femad-6

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Creating AI-generated media requires knowing which model fits each task, what parameters each endpoint accepts, and how to control cost and reproducibility. This Skill consolidates image, video, and audio generation through the fal.ai MCP server into one guided workflow. ## Core Features & Use Cases - Image Generation: Text-to-image and image editing with Nano Banana 2 for fast drafts and Nano Banana Pro for high-fidelity output, with control over aspect ratio, seed, and guidance scale. - Video Generation: Text-to-video and image-to-video via Seedance 1.0 Pro, Kling Video v3 Pro, and Veo 3, including duration and aspect ratio parameters. - Audio Generation: Conversational text-to-speech with CSM-1B, video-to-audio with ThinkSound, plus ElevenLabs API and VideoDB options for voice, music, and sound effects. - Use Case: A content creator needs a thumbnail, a short promo clip, and a voiceover. They iterate on the thumbnail with Nano Banana 2, generate the clip from the final image with Seedance, and produce narration with CSM-1B, checking estimate_cost before each expensive run. ## Quick Start Configure the fal.ai MCP server with your FAL_KEY, then ask the assistant to generate an image of a specific scene using the fal-ai nano-banana-2 model.

Frequently Asked Questions about fal-ai-media

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I generate images with fal.ai?

Use the generate tool with app_id fal-ai/nano-banana-2 for fast drafts or fal-ai/nano-banana-pro for high-fidelity output. Pass a prompt, image_size such as landscape_16_9, num_images, and optionally a seed for reproducible results.

How do I create a video from an image with fal.ai?

Upload the source image with the upload tool, then call generate on fal-ai/seedance-1-0-pro with the image_url, a motion-focused prompt, and a duration of 5s or 10s. Image-to-video gives more controlled results than pure text-to-video.

Which fal.ai video model should I use, Seedance, Kling, or Veo 3?

Seedance 1.0 Pro suits text-to-video and image-to-video with high motion quality. Kling Video v3 Pro adds native audio generation, and Veo 3 produces video with generated sound and high visual quality.

Does fal.ai media generation require an API key?

Yes, the fal.ai MCP server must be configured with a FAL_KEY environment variable obtained from fal.ai. The server is registered in the MCP configuration using npx fal-ai-mcp-server.

How do I estimate fal.ai generation cost before running a model?

Call the estimate_cost tool with estimate_type unit_price and the target endpoint such as fal-ai/nano-banana-pro with its unit_quantity. This lets you check pricing before running expensive video generations.

Can I generate speech and sound effects without the fal.ai MCP?

Yes, the Skill documents alternatives: call the ElevenLabs text-to-speech API directly with an API key, or use VideoDB generative audio for voice, music, and sound effects if VideoDB is configured.