SporeLabs

Your LLM bill goes down every week.
You do nothing.

Change one line and fund the wallet. We test cheaper models against your traffic. A call moves only when the cheaper model matches.

OpenAI-compatible Prepaid wallet No markup on upstream spend

app/llm.py
from openai import OpenAI
client = OpenAI(
base_url="https://api.openai.com/v1",
api_key=os.environ["LLM_KEY"],
)
That is the whole change. Keep sending model: "gpt-4o". A named model is honored for that call.

Switch my LLM calls to SporeLabs — set the OpenAI base URL to https://sporelabs.dev/v1 and read the key from SPORELABS_API_KEY. Change nothing else.

Hand it to your coding agent or connect the application yourself.

Example

Every week, the bill goes down.

Week 0

Example: monthly cost falling over twelve weeks

Without us: $840/mo

$840/mo

Off the bill

$0

Reading

    Example at published Sonnet rates. Yours will differ.

    How small this can get

    The bot that answers you here costs us nothing to run.

    A 1.7-billion-parameter model downloads into this tab and answers on your hardware. No server, no key, no request to us.

    Qwen3 1.7B · on your device

    Ask it anything

    The first message downloads the weights. Everything after that is instant and free.

    One line now. A cheaper system every week after.

    Put it to work Set up your coding agent support@sporelabs.dev