The Fastest InferencePowered by Autonomous Research
gpt-oss-120b
OpenAI's open-weight reasoning model, served on our optimized stack.
Input
$0.15
Output
$0.60
per M tokens
gpt-oss-120b-fast
Flagship low-latency tier — the fastest gpt-oss on GPU.
Input
$0.30
Output
$1.20
per M tokens
GLM 5.2
GLM 5.2Z.ai's flagship coding and reasoning model.
Input
$1.00
Output
$4.00
per M tokens
World-Class Autonomous Research
Read our work on creating the autonomous research systems that created this API.
Blazingly Fast Inference
We coordinate populations of research agents to create the fastest gpt-oss-120b deployment on GPU.
Fastest API providers output speed on GPU
for gpt-oss-120b
We built Aster to automate research, and quickly learned that the limiting reagent of autonomous science is inference. So we used our system on inference itself: Aster optimizes the stack it runs on, and every improvement compounds into more research per dollar.
Aster Inference is that same stack, offered as an API — and it keeps getting faster and cheaper as Aster improves it.
FAQs
Do you retain my data?
No — zero data retention by default. Prompts and outputs are processed in memory and never stored. We keep only token counts for billing.
How do I get access?
Create an API key in the console and add credits. For volume pricing, contact info@asterlab.ai.
Which SDKs and tools work with it?
Anything that speaks the OpenAI API — the official OpenAI Python and JS SDKs, LangChain, LiteLLM, Cursor, the Vercel AI SDK, and most agent frameworks. Set the base URL to https://api.asterlab.ai/v1 and pass your Aster key.
Why is it so fast?
Our autonomous research system continuously optimizes the stack it runs on.