Serverless guide
Serverless prompt compression
Serverless functions have tight resource limits. SuperCompress adds prompt compression in ~60ms with no GPU, no model downloads, and minimal memory — perfect for serverless deployments.
AWS Lambda deployment
# Lambda function that compresses before calling an LLM
import json
from supercompress import Compressor
comp = Compressor()
def lambda_handler(event, context):
body = json.loads(event["body"])
result = comp.compress(body["context"], body["query"])
# Forward compressed context to your LLM
return {
"statusCode": 200,
"body": json.dumps({
"compressed": result.compressed_text,
"savings": result.tokens_removed
})
}Serverless compatibility
| Platform | Cold Start | Memory | Compression Time |
|---|---|---|---|
| AWS Lambda | ~300ms | ~80MB | ~60ms |
| Google Cloud Functions | ~200ms | ~80MB | ~60ms |
| Cloudflare Workers | ~5ms | ~50MB | ~70ms |
| Vercel Edge Functions | ~50ms | ~60MB | ~65ms |
Frequently asked questions
Does SuperCompress fit in Lambda's /tmp space?
Yes. The package is ~200KB. No model files needed.
Can I use it in edge runtimes?
Yes. The compressor is pure Python and works in edge environments that support Python.
Build with less context
Put compression in front of your next LLM call.
Use the hosted API or run SuperCompress locally. Keep the evidence, drop the token waste, and measure the savings before it reaches your model.