OLLAMA IN THE CLOUD
Your Ollama models, cloud-powered.
Love Ollama but limited by your local GPU? Run the same models in the cloud through Runcrate's API. Llama, Qwen, DeepSeek, Gemma, Mistral, and more, all accessible via an OpenAI-compatible endpoint. No VRAM constraints, no thermal throttling, no model downloads. Just swap the endpoint.
QUICK START
Integrate in minutes.
from openai import OpenAI
# Same code you'd use with Ollama's OpenAI-compatible mode
client = OpenAI(
base_url="https://api.runcrate.ai/v1",
api_key="rc_live_YOUR_API_KEY",
)
response = client.chat.completions.create(
model="meta-llama/Llama-4-Scout-17B-16E-Instruct",
messages=[
{"role": "user", "content": "Summarize the benefits of cloud inference."}
],
)
print(response.choices[0].message.content)AVAILABLE MODELS
Models you can use today.
| Model | Provider | Price | Detail |
|---|---|---|---|
| meta-llama/Llama-4-Scout-17B-16E-Instruct | Meta | Per-token | 17B MoE, 128K context |
| Qwen/Qwen3-32B | Alibaba | Per-token | 32B, strong multilingual |
| deepseek-ai/DeepSeek-V3 | DeepSeek | Per-token | 128K context, MoE |
| google/gemma-3-27b-it | Per-token | 27B, instruction-tuned | |
| mistralai/Mistral-Small-24B-Instruct-2501 | Mistral | Per-token | 24B, fast inference |
WHY RUNCRATE
Built for production.
No Local GPU Needed
Stop waiting for model downloads and fighting with CUDA versions. Your models run on H100s in the cloud, accessible from anywhere.
Same Models, Bigger Scale
The models you know from Ollama, running on enterprise-grade hardware. No VRAM limits, no thermal throttling, no 4-bit quantization compromises.
OpenAI-Compatible API
If your code works with Ollama's OpenAI compatibility mode, it works with Runcrate. Change the base URL and API key, done.
Pay Per Token
No idle GPU costs. You pay for the tokens you generate, not for a machine sitting warm. Ideal for bursty or unpredictable workloads.
COMPARISON
Runcrate vs Local Ollama.
| Feature | Runcrate | Local Ollama |
|---|---|---|
| GPU hardware | H100 / H200 cloud | Your local GPU |
| Model size limit | No VRAM limit | Limited by your GPU |
| Setup | API key + one line | Download + install |
| Cost model | Per-token, no idle cost | Electricity + GPU purchase |
| Available models | 200+ via API | Community-maintained |
FAQ