modal-serverless-gpu
About
Modal provides a serverless GPU cloud platform for running ML workloads without infrastructure management. It enables deploying models as auto-scaling APIs and running batch jobs with pay-per-second pricing. Key features include on-demand access to various GPUs (T4 to H100) and a Python-native interface for defining compute tasks.
Quick Install
Claude Code
Recommended/plugin add https://github.com/davila7/claude-code-templatesgit clone https://github.com/davila7/claude-code-templates.git ~/.claude/skills/modal-serverless-gpuCopy and paste this command in Claude Code to install this skill
Documentation
Modal Serverless GPU
Comprehensive guide to running ML workloads on Modal's serverless GPU cloud platform.
When to use Modal
Use Modal when:
- Running GPU-intensive ML workloads without managing infrastructure
- Deploying ML models as auto-scaling APIs
- Running batch processing jobs (training, inference, data processing)
- Need pay-per-second GPU pricing without idle costs
- Prototyping ML applications quickly
- Running scheduled jobs (cron-like workloads)
Key features:
- Serverless GPUs: T4, L4, A10G, L40S, A100, H100, H200, B200 on-demand
- Python-native: Define infrastructure in Python code, no YAML
- Auto-scaling: Scale to zero, scale to 100+ GPUs instantly
- Sub-second cold starts: Rust-based infrastructure for fast container launches
- Container caching: Image layers cached for rapid iteration
- Web endpoints: Deploy functions as REST APIs with zero-downtime updates
Use alternatives instead:
- RunPod: For longer-running pods with persistent state
- Lambda Labs: For reserved GPU instances
- SkyPilot: For multi-cloud orchestration and cost optimization
- Kubernetes: For complex multi-service architectures
Quick start
Installation
pip install modal
modal setup # Opens browser for authentication
Hello World with GPU
import modal
app = modal.App("hello-gpu")
@app.function(gpu="T4")
def gpu_info():
import subprocess
return subprocess.run(["nvidia-smi"], capture_output=True, text=True).stdout
@app.local_entrypoint()
def main():
print(gpu_info.remote())
Run: modal run hello_gpu.py
Basic inference endpoint
import modal
app = modal.App("text-generation")
image = modal.Image.debian_slim().pip_install("transformers", "torch", "accelerate")
@app.cls(gpu="A10G", image=image)
class TextGenerator:
@modal.enter()
def load_model(self):
from transformers import pipeline
self.pipe = pipeline("text-generation", model="gpt2", device=0)
@modal.method()
def generate(self, prompt: str) -> str:
return self.pipe(prompt, max_length=100)[0]["generated_text"]
@app.local_entrypoint()
def main():
print(TextGenerator().generate.remote("Hello, world"))
Core concepts
Key components
| Component | Purpose |
|---|---|
App | Container for functions and resources |
Function | Serverless function with compute specs |
Cls | Class-based functions with lifecycle hooks |
Image | Container image definition |
Volume | Persistent storage for models/data |
Secret | Secure credential storage |
Execution modes
| Command | Description |
|---|---|
modal run script.py | Execute and exit |
modal serve script.py | Development with live reload |
modal deploy script.py | Persistent cloud deployment |
GPU configuration
Available GPUs
| GPU | VRAM | Best For |
|---|---|---|
T4 | 16GB | Budget inference, small models |
L4 | 24GB | Inference, Ada Lovelace arch |
A10G | 24GB | Training/inference, 3.3x faster than T4 |
L40S | 48GB | Recommended for inference (best cost/perf) |
A100-40GB | 40GB | Large model training |
A100-80GB | 80GB | Very large models |
H100 | 80GB | Fastest, FP8 + Transformer Engine |
H200 | 141GB | Auto-upgrade from H100, 4.8TB/s bandwidth |
B200 | Latest | Blackwell architecture |
GPU specification patterns
# Single GPU
@app.function(gpu="A100")
# Specific memory variant
@app.function(gpu="A100-80GB")
# Multiple GPUs (up to 8)
@app.function(gpu="H100:4")
# GPU with fallbacks
@app.function(gpu=["H100", "A100", "L40S"])
# Any available GPU
@app.function(gpu="any")
Container images
# Basic image with pip
image = modal.Image.debian_slim(python_version="3.11").pip_install(
"torch==2.1.0", "transformers==4.36.0", "accelerate"
)
# From CUDA base
image = modal.Image.from_registry(
"nvidia/cuda:12.1.0-cudnn8-devel-ubuntu22.04",
add_python="3.11"
).pip_install("torch", "transformers")
# With system packages
image = modal.Image.debian_slim().apt_install("git", "ffmpeg").pip_install("whisper")
Persistent storage
volume = modal.Volume.from_name("model-cache", create_if_missing=True)
@app.function(gpu="A10G", volumes={"/models": volume})
def load_model():
import os
model_path = "/models/llama-7b"
if not os.path.exists(model_path):
model = download_model()
model.save_pretrained(model_path)
volume.commit() # Persist changes
return load_from_path(model_path)
Web endpoints
FastAPI endpoint decorator
@app.function()
@modal.fastapi_endpoint(method="POST")
def predict(text: str) -> dict:
return {"result": model.predict(text)}
Full ASGI app
from fastapi import FastAPI
web_app = FastAPI()
@web_app.post("/predict")
async def predict(text: str):
return {"result": await model.predict.remote.aio(text)}
@app.function()
@modal.asgi_app()
def fastapi_app():
return web_app
Web endpoint types
| Decorator | Use Case |
|---|---|
@modal.fastapi_endpoint() | Simple function → API |
@modal.asgi_app() | Full FastAPI/Starlette apps |
@modal.wsgi_app() | Django/Flask apps |
@modal.web_server(port) | Arbitrary HTTP servers |
Dynamic batching
@app.function()
@modal.batched(max_batch_size=32, wait_ms=100)
async def batch_predict(inputs: list[str]) -> list[dict]:
# Inputs automatically batched
return model.batch_predict(inputs)
Secrets management
# Create secret
modal secret create huggingface HF_TOKEN=hf_xxx
@app.function(secrets=[modal.Secret.from_name("huggingface")])
def download_model():
import os
token = os.environ["HF_TOKEN"]
Scheduling
@app.function(schedule=modal.Cron("0 0 * * *")) # Daily midnight
def daily_job():
pass
@app.function(schedule=modal.Period(hours=1))
def hourly_job():
pass
Performance optimization
Cold start mitigation
@app.function(
container_idle_timeout=300, # Keep warm 5 min
allow_concurrent_inputs=10, # Handle concurrent requests
)
def inference():
pass
Model loading best practices
@app.cls(gpu="A100")
class Model:
@modal.enter() # Run once at container start
def load(self):
self.model = load_model() # Load during warm-up
@modal.method()
def predict(self, x):
return self.model(x)
Parallel processing
@app.function()
def process_item(item):
return expensive_computation(item)
@app.function()
def run_parallel():
items = list(range(1000))
# Fan out to parallel containers
results = list(process_item.map(items))
return results
Common configuration
@app.function(
gpu="A100",
memory=32768, # 32GB RAM
cpu=4, # 4 CPU cores
timeout=3600, # 1 hour max
container_idle_timeout=120,# Keep warm 2 min
retries=3, # Retry on failure
concurrency_limit=10, # Max concurrent containers
)
def my_function():
pass
Debugging
# Test locally
if __name__ == "__main__":
result = my_function.local()
# View logs
# modal app logs my-app
Common issues
| Issue | Solution |
|---|---|
| Cold start latency | Increase container_idle_timeout, use @modal.enter() |
| GPU OOM | Use larger GPU (A100-80GB), enable gradient checkpointing |
| Image build fails | Pin dependency versions, check CUDA compatibility |
| Timeout errors | Increase timeout, add checkpointing |
References
- Advanced Usage - Multi-GPU, distributed training, cost optimization
- Troubleshooting - Common issues and solutions
Resources
- Documentation: https://modal.com/docs
- Examples: https://github.com/modal-labs/modal-examples
- Pricing: https://modal.com/pricing
- Discord: https://discord.gg/modal
GitHub Repository
Related Skills
railway-deployment
MetaThis Claude Skill manages Railway deployments for lifecycle operations and troubleshooting. It enables developers to view logs, redeploy, restart, or remove deployments through Railway CLI commands. Use it for deployment visibility and debugging, but note that deleting services requires the railway-environment skill instead.
railway-database
MetaThis skill adds official Railway database services (Postgres, Redis, MySQL, MongoDB) with pre-configured volumes and connection variables. Use it when developers request to add, connect, or wire up databases in their Railway projects. It specifically handles database services while directing other templates to the separate railway-templates skill.
railway-status
MetaThis skill checks the current deployment status and uptime of Railway projects in the current directory. It's triggered by queries like "railway status," "what's deployed," or questions about deployment status and uptime. Use the separate railway-environment skill for configuration or variable queries instead.
railway-templates
MetaThis skill enables developers to search for and deploy pre-configured services from Railway's template marketplace, such as Ghost, Strapi, and n8n. Use it when you need to quickly add a templated service or find templates for a specific use case like CMS or monitoring. For core databases, the separate railway-database skill is preferred.
