fal-model-guide

## Quick Reference

| Category | Fast/Cheap | Balanced | Best Quality |
|----------|------------|----------|--------------|
| Image | FLUX Schnell | FLUX.1 Dev | FLUX.2 Pro |
| Video | Runway Turbo | LTX-2 Pro | Kling 2.6 Pro |
| Audio STT | Whisper Turbo | Whisper | Whisper Large v3 |
| Audio TTS | Kokoro | XTTS | ElevenLabs |

| FLUX Model | Endpoint | Steps | Use Case |
|------------|----------|-------|----------|
| FLUX.2 Pro | `fal-ai/flux-2-pro` | 28 | Production |
| FLUX.1 Dev | `fal-ai/flux/dev` | 28 | High quality |
| FLUX Schnell | `fal-ai/flux/schnell` | 4 | Fast iteration |

| Video Model | Duration | Audio | Speed |
|-------------|----------|-------|-------|
| Kling 2.6 Pro | 5-10s | Yes | Slow |
| LTX-2 Pro | 5s | Yes | Medium |
| Runway Turbo | 5-10s | No | Fast |

## When to Use This Skill

Use for **model selection decisions**:
- Comparing FLUX vs SDXL for image generation
- Choosing video models by quality tier
- Optimizing costs with appropriate model selection
- Understanding model-specific parameters
- Building production vs development pipelines

**Related skills:**
- For text-to-image: see `fal-text-to-image`
- For text-to-video: see `fal-text-to-video`
- For audio models: see `fal-audio`

---

# fal.ai Model Selection Guide

Comprehensive guide to selecting the right fal.ai model for your use case.

## Image Generation Models

### FLUX Family

| Model | Endpoint | Best For | Speed | Quality | Cost |
|-------|----------|----------|-------|---------|------|
| FLUX.2 [pro] | `fal-ai/flux-2-pro` | Production, best quality | Medium | Highest | $$$ |
| FLUX.1 [dev] | `fal-ai/flux/dev` | High quality, open-source | Medium | High | $$ |
| FLUX Schnell | `fal-ai/flux/schnell` | Fast iteration | Fast | Good | $ |
| FLUX LoRA | `fal-ai/flux-lora` | Custom trained styles | Medium | High | $$ |
| FLUX Realism | `fal-ai/flux-realism` | Photorealistic images | Medium | High | $$ |

**FLUX.2 Pro** - Latest and best quality
```typescript
const result = await