OLLAMA IN THE CLOUD
Your Ollama models, cloud-powered.
Run the models you use in Ollama without a local GPU. Llama, Qwen, DeepSeek, Gemma and Mistral through an OpenAI-compatible endpoint: change the base URL and keep going.
Hand this to your agent
Add a Runcrate model to this project: run curl -fsSL"https://runcrate.ai/agents.md?intent=build"and follow it.
Your agent asks for an API key. Create a key
Quick start
Hand it to your agent. Or read the request it sends.
Your agent writes and runs it. The same key works for every model and every GPU.
Available models
Models you can call today. Billed prices, as charged.
| Model | Maker | Price | Detail |
|---|---|---|---|
| meta-llama/Llama-4-Scout-17B-16E-Instruct | Meta | $0.105 in, $0.315 out per 1M tokens | 17B MoE, 128K context |
| google/gemma-3-27b-it | $0.084 in, $0.168 out per 1M tokens | 27B, instruction-tuned |
Why Runcrate
Built for production. One key, one balance.
No local GPU needed
Stop waiting for model downloads and fighting with CUDA versions. The models run in the cloud, reachable from anywhere.
Same models, bigger scale
The models you know from Ollama, running on datacenter GPUs. No VRAM limits on your side and no thermal throttling.
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 and Local Ollama, side by side.
| Feature | Runcrate | Local Ollama |
|---|---|---|
| GPU hardware | Cloud GPUs, nothing to install | 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 | Open models via API | Community-maintained |
FAQ