Lyceum

What's good about Lyceum

• Instant availability of GPUs with minimal provisioning time.
• No capacity planning or vendor lock-in required.
• Complies with EU data residency requirements.

Lyceum GPU prices

6 GPU models across 11 configurations. Rank compares Lyceum's cheapest on-demand listing for each GPU with every provider's cheapest.

What a run costs on Lyceum

24 hours at today's cheapest listing for each GPU, at the size it is listed.

• 1× H100 SXM for 24 hours

  At $4.69/GPU-hr

  $112.56

  $30.30 over the market median rate

• 1× H200 for 24 hours

  At $6.19/GPU-hr

  $148.56

  $39.89 over the market median rate

• 1× L40S for 24 hours

  At $1.99/GPU-hr

  $47.76

  $21.60 over the market median rate

The market median is every provider's cheapest on-demand price for that GPU.

Lyceum LLM API prices

Pay-per-token pricing, per 1M tokens.

Prices last updated October 30, 2026

Billing & terms

Serverless inference

Billed per token for input, cached input and output, with no base fee

Prompt caching

Repeated prompt prefixes still in GPU memory are billed at the cached input rate where listed

On-demand GPUs

Billed per second with no subscription

Spot GPUs

About 45% less than on-demand; capacity can be reclaimed

Reserved GPUs, clusters and dedicated endpoints

Commitment pricing agreed per contract, including multi-year reserved blocks

Regions

European data centres; serverless models run in Paris and Finland, while dedicated and VM locations are agreed per contract

Support

Email support for all customers; business customers also get a shared Slack channel and agreed response times.

Watch out for

• Dedicated endpoint and reserved capacity pricing is agreed per contract rather than published.
• Spot GPUs can be reclaimed, and not every GPU type is offered as spot.
• Data centres are in Europe only.

Features

Serverless Inference

API access to open-source models with pay-per-token pricing.

Dedicated Endpoints

Reserved GPU capacity for production models ensuring low latency and high availability.

Serverless Training

Run training jobs on GPUs without the need for infrastructure management.

GPU Virtual Machines

Full root access to customizable GPU instances ready in seconds.

Large-Scale Clusters

Support for configurations from 8 to 8,000 GPUs with InfiniBand connectivity.

OpenAI-Compatible API

Serverless models are served through an OpenAI-compatible endpoint, so existing OpenAI SDK code works by changing the base URL and API key.

Zero Data Retention

Prompts and outputs are processed but not stored or written to disk, and are not used for training.

Object Storage

Per-user S3-compatible bucket for inputs, outputs, and shared datasets.

Dashboard, CLI and REST API

The same capabilities are exposed through a web dashboard, the lyceum CLI, a VS Code extension and a REST API, with SSO available for organizations.

Getting started

1. Step 1

   Sign Up

   Create an account on the Lyceum platform.

2. Step 2

   Choose a Service

   Select between inference, training, or virtual machine options.

3. Step 3

   Deploy a Model or Job

   Use provided API or CLI instructions to submit your workload.

4. Step 4

   Monitor Usage

   Utilize dashboard features to track resource usage and performance.

5. Step 5

   Seek Support if Needed

   Contact the sales or support team for assistance.