Deepfake detection and media intelligence for trust and safety teams, delivered through an API or MCP.
Contact sales
Crusoe Cloud
AI cloud with on-demand GPU instances, managed Kubernetes and Slurm, and managed inference and fine-tuning for open models.
What is Crusoe Cloud?
Crusoe Cloud is an infrastructure provider for AI training and inference. Teams rent GPU and CPU instances by the hour, with spot, on-demand and reserved pricing, and can build multi-node clusters orchestrated through Crusoe Managed Kubernetes or Managed Slurm. The GPU lineup on the pricing page includes NVIDIA Vera Rubin, B200, H200 and H100 parts and AMD MI355X and MI300X. For teams that do not want to run hardware, Crusoe Managed Inference exposes open models through an OpenAI-compatible API in the Intelligence Foundry, either as pay-per-token Serverless Inference or as Self-Serve Deployments on dedicated GPUs. Its MemoryAlloy cluster-wide KV cache is described as cutting time-to-first-token by up to 9.9x. Serverless Fine-Tuning trains LoRA adapters on open models without managing GPUs. Provisioned Throughput is sold in AI Model Units with discounts for longer commitments. Prices depend on the hardware or model, so check the Crusoe Cloud pricing page and contact sales for committed rates.
Use cases
- Rent GPU instances by the hour
- Train models on a multi-node cluster
- Serve open models through an API
- Fine-tune open models with LoRA
- Run Kubernetes or Slurm for AI jobs
- Reserve inference throughput for production
Who it's for
Why use Crusoe Cloud
- You can start with pay-per-token inference and move to dedicated or provisioned capacity as traffic grows.
- One provider covers raw GPU clusters, managed orchestration and hosted model endpoints.
- The inference API is OpenAI-compatible, so existing client code can be pointed at it with little change.
- Spot, on-demand and reserved options let you trade price against guaranteed capacity.
- Support is staffed around the clock, and the vendor cites a sub-6-minute average first response.
Crusoe Cloud features
- GPU instances with spot, on-demand and reserved pricing
- CPU instances and storage for data processing and checkpoints
- Crusoe Managed Kubernetes and Managed Slurm
- Serverless Inference with pay-per-token billing
- Self-Serve Deployments on dedicated GPUs without shared rate limits
- Provisioned Throughput sold in AI Model Units
- Serverless Fine-Tuning for LoRA adapters
- OpenAI-compatible inference API
- Crusoe Command Center, CLI and Terraform support
- Cloud MCP server for querying your infrastructure from AI assistants
Crusoe Cloud integrations
How to get started with Crusoe Cloud
- Create a Crusoe Cloud account and install the CLI or set up Terraform.
- Pick a path: provision GPU VMs or a Kubernetes or Slurm cluster, or use Managed AI products.
- For inference, choose a model in the Intelligence Foundry and generate an API key for Serverless Inference.
- For steady production traffic, move to a Self-Serve Deployment or Provisioned Throughput.
- Upload a dataset to fine-tune an open model, then serve the resulting adapter.
Crusoe Cloud limitations
- Serverless Fine-Tuning was described as private preview with access by request
- Serverless Inference applies per-minute limits on both tokens and requests
- Committed, monthly and volume rates require contacting sales
Crusoe Cloud pricing
Crusoe Cloud has no free plan.
Prices can change and may vary by country or exclude tax. Check the official pricing page.
Crusoe Cloud alternatives
The main Crusoe Cloud alternatives listed on Debuly are CoreWeave, Lambda, Together AI, Fireworks AI, Runpod and Novita AI. The table compares pricing and platforms; order is A–Z.
| Tool | Pricing | Starting price | Platforms |
|---|---|---|---|
| Crusoe Cloud | Paid | — | Web, API, CLI |
| CoreWeave | Paid | — | Web, API, Bare metal |
| Fireworks AI | Paid | $0.01, usage-based | API, Web |
| Lambda | Paid | $0.69, usage-based | Web, API, CLI |
| Novita AI | Paid | $0.61, usage-based | Web, API |
| Runpod | Paid | — | Web, API |
| Together AI | Paid | — | Web, API |
CoreWeave
AI cloud that rents NVIDIA GPU capacity by the hour, with on-demand, spot and reserved plans for model training and inference.
Fireworks AI
Inference and training platform for open models: pay-per-token serverless APIs, GPU deployments, fine-tuning and reserved capacity.
Lambda
GPU cloud for AI with on-demand NVIDIA instances and reserved 1-Click Clusters for training, fine-tuning and inference.
Novita AI
AI cloud with pay-per-token model APIs for 200+ models, dedicated GPU endpoints, agent sandboxes and GPU rentals.
Runpod
Pay-per-second GPU cloud for AI: on-demand Pods, Serverless endpoints, multi-node Clusters and ready-made public model APIs.
Together AI
AI cloud for running open-source models through serverless and batch inference, dedicated endpoints, GPU clusters and fine-tuning.
Crusoe Cloud FAQ
How much does Crusoe Cloud cost?
Pricing varies by product. GPU instances are billed by the hour, Serverless Inference per million tokens, and Self-Serve Deployments per GPU type. See the Crusoe Cloud pricing page for current rates.
Does Crusoe Cloud have a free plan?
No free plan is described on the pricing page. Consumption is pay-as-you-go, with committed rates available by contacting sales.
What pricing options does Crusoe Cloud offer?
GPU and CPU capacity is offered as spot, on-demand or reserved. Managed inference is pay-per-token through Serverless Inference, or reserved through Self-Serve Deployments and Provisioned Throughput.
Does Crusoe Cloud have an API?
Yes. Serverless Inference uses an OpenAI-compatible API, and there is also a CLI, Terraform support and a Cloud MCP server for assistants such as Claude Desktop.
Which models can I run on Crusoe Managed Inference?
The pricing page lists open models such as Llama 3.3 70B Instruct, GPT-OSS 120B, Qwen3 235B A22B Instruct and DeepSeek V4 Flash. You can also bring your own fine-tuned models.
Who is Crusoe Cloud for?
AI builders who need GPU clusters for training, app developers who want hosted open-model endpoints, and enterprises that need committed capacity and support.
What are alternatives to Crusoe Cloud?
CoreWeave, Lambda, Together AI, Fireworks AI and the major hyperscalers offer comparable GPU cloud or hosted inference services.
Sources & verification
Facts on this page come from the official sources below, each with the date we read it. Written by Debuly; the vendor did not review this page.
| Source | Type | Accessed |
|---|---|---|
| crusoe.ai/ | official | |
| crusoe.ai/cloud/pricing | pricing | |
| docs.cloud.crusoe.ai/llms.txt | docs | |
| crusoe.ai/cloud | official | |
| crusoe.ai/cloud/solutions/model-inference | official | |
| crusoe.ai/cloud/managed-inference | official |
Image credits
- Logo: from cdn.prod.website-files.com · fetched · Vendor trademark, used only to identify the product; removed on request
- Cover: none collected yet, so a generated tile is shown.
Spot an error? Report an issue with this listing. We also remove a logo or image on the owner's request.
More in Developer platforms 4
All developer platforms toolsLocal-first AI memory that keeps your conversations as verbatim text and finds them again with semantic search through MCP.
Open sourcePython, TypeScript and REST SDK for building coding agents that run locally or in Docker and Kubernetes workspaces
Open sourceAn AI infrastructure agent in Pulumi Cloud that answers questions, previews changes and opens pull requests against your IaC code.
Paid
