> ## Documentation Index
> Fetch the complete documentation index at: https://infercrane.mintlify.site/llms.txt
> Use this file to discover all available pages before exploring further.

# SGLang

> The second Runtime Contract V1 engine profile.

# SGLang

SGLang is InferCrane's second registered inference engine. The runtime profile pins the official
`lmsysorg/sglang:v0.5.12` multi-platform manifest by digest and launches its OpenAI-compatible
server through the same provider-neutral workload contract used by custom OCI images.

The image tag and launch form follow the [official SGLang release and Docker
documentation](https://github.com/sgl-project/sglang/releases/tag/v0.5.12). InferCrane resolved the
manifest digest during implementation; the immutable value is visible through
`infercrane integrations --output json` and the persisted revision.

```yaml theme={"theme":{"light":"github-light-default","dark":"vesper"}}
name: qwen-sglang
model:
  id: Qwen/Qwen3-8B
runtime:
  engine: sglang
  version: 0.5.12
compute:
  mode: elastic
provider:
  cloud: aws
  region: eu-central-1
resources:
  gpu: L40S
scaling:
  min_replicas: 1
  max_replicas: 1
routing:
  strategy: round-robin
```

The built-in profile declares the immutable image and argv, so normal users do not repeat them.
It is simulated-qualified for readiness/model identity, buffered and streaming OpenAI transport,
client cancellation, connection draining, the metrics endpoint and immutable launch intent.

Real SGLang GPU behavior—including model compatibility, tool calling, structured output,
performance and provider shutdown behavior—remains deferred until consolidated manual
qualification. Autoscaling is rejected for SGLang because normalized SGLang scaling
signals are not yet qualified.