Template Marketplace
Community-contributed deployment templates for AI inference. Share and discover optimized cloud-init scripts for vLLM, TGI, SGLang, and more.
Featured Templates
Sorted by ratingLlama 3.1 70B — vLLM
vllmOptimized single-node A100×8 deployment with tensor parallelism and PagedAttention.
★ 4.8 / 51,240 downloads
Deploy this template →Mistral 7B — TGI
tgiLightweight TGI setup for cost-efficient H100 inference with flash attention.
★ 4.6 / 5890 downloads
Deploy this template →DeepSeek-R1 — SGLang
sglangReasoning model setup with extended context and KV cache optimization.
★ 4.5 / 5560 downloads
Deploy this template →Qwen2.5-72B — vLLM
vllmMulti-lingual deployment with speculative decoding and continuous batching.
★ 4.3 / 5430 downloads
Deploy this template →Templates are community-contributed. Always review cloud-init scripts before running on your infrastructure.