For indie developers
Add open-model access without building another billing stack.
Use one OpenAI-compatible request shape and one prepaid crypto balance. MiniRouter routes available open models and returns the serving provider and request cost.
$5 minimum. Credits never expire. Deposits cannot be withdrawn.
from openai import OpenAI
client = OpenAI(
base_url="https://api.minirouter.sh/v1",
api_key="mr-live-…",
)
client.chat.completions.create(
model="meta/llama-3.3-70b",
messages=[{"role": "user", "content": "ping"}],
)Why it fits
Ship one integration and keep model choice open.
Your application uses one request format. Model IDs come from the public catalog, while the response tells you which provider served the call and what it cost.
- API shape
- OpenAI-compatible chat completions, completions, embeddings, and model listing.
- Billing
- One prepaid balance covers the approved public catalog.
- Limits
- Per-key rate and spend limits are available for unattended workloads.
From key to first call
Three things to do.
- 01
Issue a key
Save the API key and permanent dashboard link before they leave the screen.
- 02
Fund it once
Add at least $5 in crypto. Credits remain until your app spends them.
- 03
Use the published API
Send a normal OpenAI-compatible request with a model ID from the catalog.
Before you connect
Know where this offer stops.
- Prepaid only
- The accountless path stops when the balance reaches zero.
- No withdrawals
- Deposits buy inference credit and cannot be returned as cash.
- No frontier inventory yet
- Restricted models stay unpublished until their resale terms are approved.
Built by SolWorks Labs
The team behind shipped products and public Solana tools including Poll, Eject, SolDisperse, SolToolkit, and IDL Type Generator. See SolWorks Labs →