Recursive x402 Agents
Delegate work. Pay for inference as agents run.
Explore the ExampleBuild with Katara
Bring inference, agent execution, and authority into your application. Start with the piece you need and connect the others as your workflow grows.
For Developers
Give your agents direct access to Katara Cortex inference. Pay per request in USDC with x402, without an account or API key.
Build applications with an API key and the OpenAI-compatible API.
Katara Cortex handles provider selection, routing, validation, metering, and settlement. Add Katara Synapse execution and Katara Locus governance as your workflow grows.
Machine-to-Machine Inference
POSThttps://api.katara.com/v1/chat/completionscurl -i https://api.katara.com/v1/chat/completions \
-H 'Content-Type: application/json' \
-d '{
"model": "katara/cortex@1",
"messages": [{"role": "user", "content": "hi"}],
"max_tokens": 8
}'This unpaid request returns 402 Payment Required with a quote. Sign the payment, then resend the same body with PAYMENT-SIGNATURE to run inference.
from openai import OpenAI
import os
client = OpenAI(
base_url="https://api.katara.com/v1",
api_key=os.environ["KATARA_API_KEY"],
)
reply = client.chat.completions.create(
model="katara/cortex@1",
messages=[{"role": "user", "content": "Hello, Katara!"}],
)Build apps with the OpenAI client you already use. API-key requests pay from your Katara account balance.
Read the API-Key QuickstartDelegate work. Pay for inference as agents run.
Explore the ExampleAgent execution, orchestration, and memory.
Explore Agent ExecutionTool access, budgets, policies, and approvals.
Explore Agent GovernanceMore from Katara. Stay tuned.
For x402 Agents
An agent can delegate work to new agents, which can spawn agents of their own. Each can call Katara Cortex for inference and pay on the spot with x402.
Define a task, budget, and spending rules for each agent. It can buy inference as the workflow runs, paying in USDC for the work it needs to complete.
Recursive Agents.
Inference Paid on Demand.
POST /v1/chat/completions402 Payment RequiredPAYMENT-SIGNATURE200 OKExecution and Governance
Use Katara Synapse and Katara Locus independently or together. Your application can use Katara Cortex for inference without running a managed agent.
Define agents, connect approved knowledge and memory, and run, schedule, inspect, and evaluate their work.
Read the Execution GuideSet tool permissions, grants, budgets, policies, and required human approvals. Govern agents hosted on Katara or in your existing runtime.
Govern an External AgentComing Soon
A new part of the Katara developer experience is on its way. Follow the Katara blog for future announcements.
Follow Katara Updates