Pay-per-token API access to open-source models.
OpenAI-compatible APIs. No infrastructure to manage. Scale instantly from zero to thousands of requests per second.
No infrastructure management
Send requests, get responses. We handle everything else.
Instant scaling
From zero to thousands of requests per second, automatically.
Pay per token
No idle costs. You pay only for the tokens you process.
OpenAI & Anthropic-compatible
Drop-in replacement. Change one line of code. Works with Claude Code too.
Get started in minutes
Use the OpenAI Python library with just one line changed. Your existing code works out of the box.
from openai import OpenAI
client = OpenAI(
base_url="https://api.lyceum.technology/v1",
api_key="your-api-key"
)
response = client.chat.completions.create(
model="meta-llama/Llama-3.3-70B-Instruct",
messages=[{"role": "user", "content": "Hello!"}]
)
print(response.choices[0].message.content)Claude Code, powered by Lyceum
Run Claude Code on any of our models and switch between them mid-session with /model. One command to set up, no restart to switch.
# One-time setup
pip install lyceum-cli
lyceum auth login --api-key lk_your_key
# From then on
lyceum code
# Select Yes at the API key prompt on first start
# Switch models mid-session, no restart needed
/model z-ai/glm-5.2
/model moonshotai/kimi-k2.6
/model deepseek/deepseek-v4-proUsing Cursor? Point its OpenAI base URL at Lyceum to chat with our models in the editor. Cursor setup guide →
Supported models
Access the latest open-source models through a single API.
Kimi K2.7 Code
MoonshotMiniMax M3
MiniMaxDeepSeek V4 Pro
DeepSeekDeepSeek V4 Flash
DeepSeekGLM-5.2
ZAIKimi K2.6
MoonshotQwen3.5 397B A17B
Qwengpt-oss 120B
OpenAILlama 3.3 70B
MetaSimple, transparent pricing
Pay only for the tokens you use. No minimum spend, no hidden fees. Volume discounts available.
View pricingReady to get started?
Sign up and start sending requests in minutes.