Inference platform · general availability

Serve your own models.
We run the GPUs.

ModelCloud is the serving layer for your AI models. Bring your own weights or an open-weight model — we handle GPU inference, autoscaling, and observability, so your team ships product instead of babysitting infrastructure. No vendor lock-in, ever.

$ modelcloud deploy — takes about a minute from weights to a live endpoint
deploy — my-model · a10g · autoscale 1..8 bash
$ modelcloud login # Authenticated as priya@modelcloud.cloud $ modelcloud deploy ./llama-3-8b-instruct \     --runtime vllm --gpu a10g --autoscale 1..8 weights uploaded (6.2 GiB, 41s) model optimized · 4-bit AWQ, 128 ctx cache endpoint live → https://api.modelcloud.cloud/v1/my-model $ curl https://api.modelcloud.cloud/v1/my-model \     -d '{"prompt":"write a haiku about on-call"}'
live endpoint
status● serving
replicas1 → 8
p50 latency214 ms
cold start0.9 s
req / min▲ 412
scale-to-zeroon
How it works

You bring the model. We run it.

We're not a reseller of someone else's models. ModelCloud is the infrastructure your models run on — the runtime, the scaling, the telemetry — while you keep full ownership of your weights and your data.

01 / DEPLOY

Bring your weights

Point at a Hugging Face repo, a container, or a private registry. We pull the weights, quantize, and prep the runtime — vLLM, TensorRT-LLM, or SGLang.

modelcloud deploy
02 / SCALE

Autoscale on demand

Set a min and max replica range. ModelCloud scales on request volume or queue depth, and scales to zero when traffic stops — you pay only for what you run.

--autoscale 1..8
03 / OBSERVE

Watch it run

Token-level latency, GPU utilization, queue depth, and traces per request. Export to your own metrics stack — the data is yours, not locked in ours.

modelcloud logs -f
The platform

Everything between your weights and a production endpoint.

GPU inference runtimes

vLLM, TensorRT-LLM, SGLang, and llama.cpp for CPU — pick the engine that fits your latency and throughput targets. We tune the serving parameters so you don't have to.

Cold starts in seconds

Weights stay warm on a preloaded cache, so scale-from-zero doesn't mean a multi-minute stall. Warm containers resume in under a second; cold pulls are optimized and pipelined.

Multi-GPU sharding

Tensor-parallel and pipeline-parallel sharding for large open models that don't fit on a single accelerator. One flag, not a config labyrinth.

Your model, your data

Weights and inference traffic stay in your account. Deploy to your own VPC, bring your own cloud, or run fully managed. There's no shared multi-tenant wall behind your requests.

Developer experience

An API that feels native, not bolted on.

One OpenAI-compatible endpoint for every model you deploy, with streaming, function calling, and structured output where the runtime supports it.

  • Drop-in /v1/chat/completions compatibility
  • Streaming, batching, and priority queues
  • Per-request traces and token accounting
  • CLI, Python, and TypeScript SDKs
# stream from your own model
import modelcloud

client = modelcloud.Client()

resp = client.chat.completions.create(
    model="llama-3-8b",
    messages=[{"role": "user",
               "content": "Summarize this log"}],
    stream=True,
)
for chunk in resp:
    print(chunk.choices[0].delta.content)
What builders say

Loved by teams who run their own models.

★★★★★

"We moved our fine-tuned code model off a hand-rolled k8s setup. Cold starts went from 'go get coffee' to under a second, and we deleted a whole repo of YAML."

SL
Sarah Lindqvist
Staff ML Engineer, e-commerce platform
★★★★★

"The observability is what sold me — token-level traces I could export straight into our existing Grafana. I finally know exactly where latency goes."

YT
Yuki Tanaka
Principal Engineer, developer tools
★★★★★

"Scale-to-zero changed our compute bill without me touching a single model. We serve a dozen open-weight models and pay for what actually runs."

DA
Dev Anand
Head of AI, logistics & supply chain
Changelog

What's shipped recently.

View all releases →

Put your model in production this afternoon.

Deploy your first model in minutes — open-weight or your own fine-tune. The sandbox tier is free, no credit card.

Read the quickstart →