OBJECT STORAGE + AI INFERENCE
Build with Firestarter S3-compatible storage and Pipe AI inference. Manage your data and run models through familiar APIs, with clear usage-based pricing for each.
Storage: Solana wallet · Prepaid USDC
A separate key for each application. Access scoped to the work it needs to do.
LT••••••••••••7C2A
LT••••••••••••9F4B
Bucket and prefix scoping · Revoke access from the console
See available, spent, and reserved credit together. Add credit when you need it.
Confirm each payment in your wallet.
Real controls for your storage. Explore usage, keys, and billing above.
01 / YOUR BUILDING BLOCKS
Familiar interfaces.
Room for your next idea.
Store application files, datasets, and build artifacts. Connect your wallet and add prepaid credit. Manage files in PipeBox or create scoped API keys for your S3 clients.
Add chat, reasoning, and agent workflows to your application through an OpenAI-compatible API. Choose a model and start making requests.
Get an API key from Pipe inference and set
PIPE_API_KEY before running. Requests use your
inference credit.
curl https://api.pipenetwork.ai/v1/chat/completions \
-H "Authorization: Bearer $PIPE_API_KEY" \
-H 'Content-Type: application/json' \
-d '{"model":"pipe/deepseek-v4-flash-0731",
"messages":[{"role":"user","content":"Hello, Pipe."}]}'02 / FROM CONSOLE TO CODE
Sign in, add storage credit, and put your S3 client to work.
Open the console and sign a free message to create or access your storage account.
Pay with USDC, then create an API key scoped to the buckets and files your application needs.
Set your credentials, create a bucket, and send your first object.
After funding your account and creating a key, replace the
credential placeholders. This example uses a bucket named
my-app.
# Credentials from your Firestarter console
export AWS_ACCESS_KEY_ID='<access-key>'
export AWS_SECRET_ACCESS_KEY='<secret-key>'
export AWS_DEFAULT_REGION=us-east-1
export AWS_REQUEST_CHECKSUM_CALCULATION=when_required
export AWS_RESPONSE_CHECKSUM_VALIDATION=when_required
aws configure set default.s3.addressing_style path
# Firestarter storage endpoint
export PIPE_ENDPOINT='https://api.pipedev.network'
printf 'Hello, Pipe.\n' > hello.txt
aws --endpoint-url "$PIPE_ENDPOINT" \
s3api create-bucket --bucket my-app
aws --endpoint-url "$PIPE_ENDPOINT" \
s3 cp hello.txt s3://my-app/hello.txt03 / CLEAR FROM THE FIRST BYTE
Prepaid USDC. Metered reads and writes.
Add credit directly from your storage console.
—USDC / TB
Data downloaded
—USDC / TB
Data uploaded
— replicas configured. Replication is included in the write rate.
See current rates in storage billing.
View billing & add credit1 TB = 1,000 GB. Estimate covers the reads and writes entered, including replication. Re-uploads and other billable operations can change actual usage. This is not a monthly storage-capacity quote.
THE NETWORK BEHIND THE NAME
Pipe started with a belief that content delivery could be open and distributed, bringing capacity closer to the people using it. That idea continues to shape the infrastructure we build.
04 / A FEW USEFUL ANSWERS
Get the details before connecting your application.
Talk to our teamFirestarter supports core S3 operations using path-style addressing and SigV4. AWS CLI, boto3, and other clients need the settings in our compatibility guide. Unsupported features include versioning, lifecycle rules, and managed KMS encryption. Check the full limits before migrating a workload.
Open the storage console, connect a Solana wallet, and sign in. Add prepaid USDC credit, then create your storage API keys. You’ll need enough SOL for network fees. Sign-in is a free message signature; you approve funding separately in your wallet.
Add prepaid storage credit in USDC. Reads and writes draw from that balance at the service’s current rates. The console shows available, spent, and reserved credit, plus payment history. Read about storage billing.
Yes. Tell us what you want to store, your approximate data size, and the tools you use. We can help you check compatibility and plan a first test. For AI workloads, explore Pipe inference.