JosiahSiegel/fal-model-guide
plugins/fal-ai-master/skills/fal-model-guide/SKILL.md
Complete fal.ai model selection system. PROACTIVELY activate for: (1) Choosing image generation models (FLUX, SDXL), (2) Choosing video models (Kling, Sora, LTX), (3) Choosing audio models (Whisper, ElevenLabs), (4) Model quality vs speed comparison, (5) Cost optimization by model tier, (6) 3D generation models, (7) Model-specific parameters, (8) Development vs production model selection. Provides: Model comparison tables, decision trees, pricing tiers, performance benchmarks. Ensures optimal model selection for quality, speed, and cost.
$ install --global
skillsauthnpx skillsauth add JosiahSiegel/claude-plugin-marketplace fal-model-guideInstall this skill globally with one command. Works with Claude Code, Cursor, and Windsurf.
Security Scan Results
3 of 9 scanners reported clean
Some scanners were skipped, did not run, or reported a non-clean status. Review each row below.
3
Scanners Passed
9
Scanners in report
Clean
TrivyContainer and dependency vulnerability scanner
95%
Clean
SemgrepStatic code analysis for vulnerabilities
95%
Clean
mcp-scan (Snyk)Model Context Protocol security validation
95%
Skipped
Snyk (dep)Open source security scanning
50%
Skipped
Socket.devSupply chain security analysis
50%
Skipped
VirusTotalMulti-engine malware detection
50%
Skipped
CrowdStrikeAdvanced threat intelligence
50%
Skipped
OSV-ScannerOpen Source Vulnerability database check
50%
Skipped
OWASP Dep-Check
50%
Last scanned: May 21, 2026, 8:02 AM135.4s1 file scanned
SKILL.md
- name:
- fal-model-guide
- description:
- |
- PROACTIVELY activate for:
- (1) Choosing image generation models (FLUX, SDXL), (2) Choosing video models (Kling, Sora, LTX), (3) Choosing audio models (Whisper, ElevenLabs), (4) Model quality vs speed comparison, (5) Cost optimization by model tier, (6) 3D generation models, (7) Model-specific parameters, (8) Development vs production model selection.
- Provides:
- Model comparison tables, decision trees, pricing tiers, performance benchmarks.
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
const result = await fal.subscribe("fal-ai/flux-2-pro", {
input: {
prompt: "Professional product photo of a watch",
image_size: "square_hd",
num_inference_steps: 28,
guidance_scale: 3.5
}
});
FLUX.1 Dev - Best open-source option
const result = await fal.subscribe("fal-ai/flux/dev", {
input: {
prompt: "A serene mountain landscape",
image_size: "landscape_16_9",
num_inference_steps: 28
}
});
FLUX Schnell - Fast 4-step generation
const result = await fal.subscribe("fal-ai/flux/schnell", {
input: {
prompt: "Quick concept sketch",
num_inference_steps: 4 // Optimized for 4 steps
}
});
Stable Diffusion Models
| Model | Endpoint | Best For |
|-------|----------|----------|
| Fast SDXL | fal-ai/fast-sdxl | Speed, lower cost |
| SDXL | fal-ai/stable-diffusion-xl | Classic SDXL |
| SD 1.5 | fal-ai/stable-diffusion-v15 | Legacy compatibility |
| SDXL Turbo | fal-ai/sdxl-turbo | Ultra-fast |
// Fast SDXL - Good balance of speed and quality
const result = await fal.subscribe("fal-ai/fast-sdxl", {
input: {
prompt: "A colorful abstract painting",
image_size: "square_hd",
num_inference_steps: 25
}
});
Specialized Image Models
| Model | Endpoint | Use Case |
|-------|----------|----------|
| Recraft V3 | fal-ai/recraft-v3 | Design assets, vectors |
| Ideogram | fal-ai/ideogram | Text in images |
| Playground v2.5 | fal-ai/playground-v25 | Creative/artistic |
| Kandinsky 3 | fal-ai/kandinsky-3 | Russian model |
Image-to-Image Models
| Model | Endpoint | Use Case |
|-------|----------|----------|
| FLUX i2i | fal-ai/flux/dev/image-to-image | Transform images |
| FLUX Inpaint | fal-ai/flux/dev/inpainting | Edit regions |
| FLUX ControlNet | fal-ai/flux/dev/controlnet | Guided generation |
| IP-Adapter | fal-ai/ip-adapter-flux | Style transfer |
// Image-to-Image transformation
const result = await fal.subscribe("fal-ai/flux/dev/image-to-image", {
input: {
image_url: "https://example.com/photo.jpg",
prompt: "Transform into watercolor painting style",
strength: 0.75 // 0-1, how much to change
}
});
// Inpainting (edit specific regions)
const result = await fal.subscribe("fal-ai/flux/dev/inpainting", {
input: {
image_url: "https://example.com/photo.jpg",
mask_url: "https://example.com/mask.png",
prompt: "A golden retriever"
}
});
// ControlNet (structural guidance)
const result = await fal.subscribe("fal-ai/flux/dev/controlnet", {
input: {
prompt: "Modern house design",
control_image_url: "https://example.com/edges.png",
controlnet_conditioning_scale: 0.8
}
});
Video Generation Models
Text-to-Video
| Model | Endpoint | Quality | Duration | Audio | Best For |
|-------|----------|---------|----------|-------|----------|
| Kling 2.6 Pro | fal-ai/kling-video/v2.6/pro | Highest | 5-10s | Native | Cinematic |
| Sora 2 | fal-ai/sora | Highest | 5-20s | Optional | Advanced |
| LTX-2 Pro | fal-ai/ltx-video-2-pro | High | 5s | Yes | Fast HQ |
| Runway Gen-3 | fal-ai/runway/gen3/turbo | High | 5-10s | No | Fast |
| Luma | fal-ai/luma-dream-machine | Good | 5s | No | Creative |
| CogVideoX | fal-ai/cogvideox | Good | 6s | No | Open source |
Kling 2.6 Pro - Best overall quality
const result = await fal.subscribe("fal-ai/kling-video/v2.6/pro", {
input: {
prompt: "A majestic eagle soaring over mountains at golden hour",
duration: 5,
aspect_ratio: "16:9",
negative_prompt: "blurry, distorted",
cfg_scale: 0.5
}
});
LTX-2 Pro - Fast with audio
const result = await fal.subscribe("fal-ai/ltx-video-2-pro", {
input: {
prompt: "Ocean waves crashing on rocks",
resolution: "720p",
enable_audio: true
}
});
Image-to-Video
| Model | Endpoint | Best For |
|-------|----------|----------|
| MiniMax Hailuo | fal-ai/minimax/video-01 | Image animation |
| Kling i2v | fal-ai/kling-video/v2.6/pro/image-to-video | HQ animation |
| Luma i2v | fal-ai/luma-dream-machine | Creative |
| Runway i2v | fal-ai/runway/gen3/turbo/image-to-video | Fast |
// Animate a still image
const result = await fal.subscribe("fal-ai/minimax/video-01", {
input: {
image_url: "https://example.com/portrait.jpg",
prompt: "Person slowly turns head and smiles",
prompt_optimizer: true
}
});
Video-to-Video (Editing)
// Edit/transform existing video
const result = await fal.subscribe("fal-ai/kling-video/o1", {
input: {
video_url: "https://example.com/video.mp4",
prompt: "Change to anime style"
}
});
Audio Models
Speech-to-Text
| Model | Endpoint | Best For |
|-------|----------|----------|
| Whisper | fal-ai/whisper | Accurate transcription |
| Whisper Turbo | fal-ai/whisper-turbo | Fast transcription |
const result = await fal.subscribe("fal-ai/whisper", {
input: {
audio_url: "https://example.com/speech.mp3",
task: "transcribe", // or "translate"
language: "en",
chunk_level: "segment"
}
});
console.log(result.text);
console.log(result.chunks); // With timestamps
Text-to-Speech
| Model | Endpoint | Best For |
|-------|----------|----------|
| Elevenlabs | fal-ai/elevenlabs | Premium voices |
| F5-TTS | fal-ai/f5-tts | Voice cloning |
| Kokoro | fal-ai/kokoro | Multi-language |
// Text-to-speech with Elevenlabs
const result = await fal.subscribe("fal-ai/elevenlabs", {
input: {
text: "Hello, welcome to our service.",
voice_id: "voice_id_here"
}
});
3D Generation Models
| Model | Endpoint | Use Case |
|-------|----------|----------|
| TripoSR | fal-ai/triposr | Image to 3D mesh |
| InstantMesh | fal-ai/instantmesh | Fast 3D generation |
| Stable Zero123 | fal-ai/stable-zero123 | Novel view synthesis |
// Generate 3D mesh from image
const result = await fal.subscribe("fal-ai/triposr", {
input: {
image_url: "https://example.com/object.jpg"
}
});
console.log(result.model_mesh.url); // GLB/OBJ file
Model Selection Decision Tree
What do you want to create?
├── Image
│ ├── From text only?
│ │ ├── Need best quality? → FLUX.2 Pro
│ │ ├── Need fast iteration? → FLUX Schnell
│ │ ├── Need open-source? → FLUX.1 Dev
│ │ └── Budget conscious? → Fast SDXL
│ ├── Transform existing image?
│ │ ├── Style transfer → FLUX i2i
│ │ ├── Edit specific region → FLUX Inpainting
│ │ └── Follow structure → FLUX ControlNet
│ └── Need text in image? → Ideogram
│
├── Video
│ ├── From text only?
│ │ ├── Need best quality? → Kling 2.6 Pro
│ │ ├── Need fast preview? → Runway Gen-3 Turbo
│ │ └── Need audio? → LTX-2 Pro or Kling
│ ├── Animate image? → MiniMax Hailuo
│ └── Edit video? → Kling O1
│
├── Audio
│ ├── Speech to text?
│ │ ├── Accurate → Whisper
│ │ └── Fast → Whisper Turbo
│ └── Text to speech? → Elevenlabs / F5-TTS
│
└── 3D
└── Image to 3D? → TripoSR
Performance Comparison
Image Generation Speed
| Model | ~Time (1024x1024) | |-------|-------------------| | FLUX Schnell | 1-2s | | Fast SDXL | 2-3s | | FLUX.1 Dev | 5-8s | | FLUX.2 Pro | 8-12s |
Video Generation Speed
| Model | ~Time (5s video) | |-------|------------------| | Runway Gen-3 Turbo | 30-60s | | LTX-2 Pro | 60-90s | | Kling 2.6 Pro | 120-180s |
Cost Optimization Strategies
-
Development Phase
- Use FLUX Schnell for prompt iteration
- Use Fast SDXL for quick tests
- Use Runway Gen-3 Turbo for video previews
-
Production Phase
- FLUX.2 Pro for final images
- Kling 2.6 Pro for final videos
-
Batch Processing
- Process during off-peak hours
- Use appropriate image sizes (don't upscale unnecessarily)
- Cache results by seed for reproducibility
-
Resource Efficiency
- Generate at target resolution (don't generate larger and downscale)
- Use webhooks for high-volume (avoid polling overhead)
- Implement client-side caching
Model-Specific Tips
FLUX Tips
- Guidance scale 3-4 works best
- 28 steps is optimal default
- Schnell works best with 4 steps only
- Detailed prompts produce better results
Video Tips
- Include camera movement in prompts
- Describe action/motion explicitly
- Use negative prompts to avoid artifacts
- Start with shorter duration for testing
Whisper Tips
- Provide language hint for better accuracy
- Use "translate" task for non-English to English
- chunk_level="segment" gives timestamps
Explore More Models
Visit https://fal.ai/models for the complete catalog of 600+ models including:
- Face restoration/enhancement
- Background removal
- Upscaling
- Style transfer
- OCR
- Object detection
- And many more specialized models
Related Skills
JosiahSiegel/ml-training
development
VerifiedTrustedCommunity
This skill should be used when the user asks to train, debug, scale, or improve ML models. PROACTIVELY activate for: (1) PyTorch, TensorFlow/Keras, JAX, Flax, Hugging Face Trainer/Accelerate training loops, (2) distributed training, DDP/FSDP/DeepSpeed, TPU/GPU setup, (3) mixed precision AMP/bf16, gradient accumulation, checkpointing, seeding, (4) overfitting, imbalance, loss functions, regularization, LR schedules, warmup, (5) memory optimization, gradient checkpointing, offloading, quantization-aware training. Provides: reproducible training best practices across deep learning and classical ML.
38SKILL.mdUpdated May 28, 2026
JosiahSiegel/ml-mlops
development
VerifiedTrustedCommunity
This skill should be used when the user asks to productionize, track, version, govern, monitor, or automate ML systems. PROACTIVELY activate for: (1) MLflow, Weights & Biases, Neptune, Comet, ClearML experiment tracking, (2) model registry, model versioning, artifact lineage, reproducibility, (3) Kubeflow, SageMaker Pipelines, Vertex AI Pipelines, Azure ML pipelines, Databricks workflows, (4) CI/CD, continuous training/evaluation, A/B tests, canary/shadow deployments, (5) drift detection, model monitoring, data validation, responsible AI governance. Provides: end-to-end MLOps architecture and operational safeguards.
38SKILL.mdUpdated May 28, 2026
JosiahSiegel/ml-inference-optimization
development
VerifiedTrustedCommunity
This skill should be used when the user asks to optimize, export, serve, compress, or accelerate ML inference. PROACTIVELY activate for: (1) latency, throughput, p95/p99, batching, concurrency, KV cache, memory, or cost issues, (2) quantization INT8/INT4, GPTQ, AWQ, bitsandbytes, pruning, sparsity, distillation, (3) ONNX export, ONNX Runtime, TensorRT, TorchScript, torch.compile, XLA, OpenVINO, Core ML, TFLite, (4) Triton, TorchServe, TF Serving, BentoML, Seldon, KServe configuration, (5) edge deployment, CPU/GPU/TPU/Inferentia serving. Provides: hardware-aware inference optimization and safe benchmarking.
38SKILL.mdUpdated May 28, 2026
JosiahSiegel/ml-hyperparameter-tuning
testing
VerifiedTrustedCommunity
This skill should be used when the user asks to tune hyperparameters, run sweeps, optimize search spaces, or use AutoML. PROACTIVELY activate for: (1) Optuna, Ray Tune, FLAML, AutoGluon, Hyperopt, Nevergrad, KerasTuner, W&B sweeps, (2) grid search, random search, Bayesian optimization, TPE, Gaussian processes, evolutionary search, (3) ASHA, Hyperband, successive halving, multi-fidelity optimization, population-based training, (4) learning-rate finder, batch-size search, early stopping, pruning, (5) reproducible sweep design and experiment analysis. Provides: budget-aware hyperparameter search strategy.
38SKILL.mdUpdated May 28, 2026