Open-source LLMs · OpenAI-compatible · prepaid credits
Affordable APIs for open-source LLMs
One OpenAI-compatible endpoint for the best open models — served on cheap GPUs and priced for builders. Sign up, get free starter credits, and ship.
from openai import OpenAI
client = OpenAI(
api_key="sk-crustoff-...",
base_url="https://api.crustoff.app/v1",
)
client.chat.completions.create(
model="qwen2.5-7b-instruct",
messages=[{"role": "user", "content": "hello"}],
)OpenAI-compatible
Drop-in /v1/chat/completions with streaming. Point any OpenAI SDK at our base URL — no rewrites.
Open models, fair prices
We run open-source LLMs ourselves on cheap GPUs and pass the savings on. Pay per token, prepaid.
No lock-in
Standard models, standard API, transparent per-token pricing. Bring your workload, leave anytime.
Pay per token. No subscriptions.
Prepaid credits. New accounts start with free credits to try it out.
Qwen2.5 7B Instruct
per 1M in / out tokens
$0.20 / $0.60
Llama 3.1 8B
permissively licensed
soon
Mixtral 8x7B
MoE throughput
soon