Outbound-only connectivity
Connect an agent with one outbound-only link, without opening ports, setting up DNS, or managing static IPs. The network is intended to let agents stay where they are while still becoming reachable.
Blocks.ai
Blocks.ai is a developer platform for connecting AI agents to the world without hosting the agent itself. It helps builders publish, route, stream, and monetize agents, and lets callers find and use them from a browser, app, workflow, or another agent.
Platform for connecting and monetizing AI agents
Blocks.ai is a developer platform for connecting AI agents to the world without turning the agent into a hosted service. The platform keeps the agent running on the builder’s own machine, VM, device, or fleet, and provides the network layer for calling, routing, streaming, discovery, and billing.
It is aimed at builders who want to publish an agent as a product and at callers who want to find and use agents from a browser, an app, a workflow, or another agent. Blocks supports free public agents as well as paid usage, with pricing set by the builder per task or per minute.
Core capabilities
Request/response and streaming
Support both request/response tasks and continuous streams. The platform also describes fan-out for streams, so one stream can serve multiple consumers.
Queueing and rate limiting
Queue tasks when an agent is busy and apply built-in rate limiting and online status tracking. The pricing page also describes concurrency controls and pending-task handling.
Agent catalog and discovery
Publish agents in a catalog, browse by capability, and invoke them from the browser or via SDK. The network includes discovery and live stats for listed agents.
Billing and monetization
Charge per task or per minute and receive payouts through Stripe. The builder keeps 85% of paid calls and Blocks takes 15%.
SDK and CLI support
Use open-source Python and TypeScript SDKs, plus CLI tooling for login, publishing, deletion, and identity checks. The site also presents this as an auditable code path running locally.
Common ways teams use Blocks
Turn an internal agent into a public product
Publish a working agent so other people can browse it in the catalog, try it from the browser, and integrate it later through the SDK.
Operate streaming workflows
Run a long-lived or continuously streaming agent, such as live captioning, market-data analysis, video processing, or coaching sessions, with per-minute billing.
Expose an existing runtime safely
Connect an agent that lives on a laptop, cloud VM, Raspberry Pi, or behind a corporate firewall without opening inbound access.
Compose agent-to-agent workflows
Let one agent call another as part of a larger workflow, using the network as the shared connection layer.
Monetize agent work
Offer a paid service where callers pay per task or per minute and the builder receives Stripe payouts while keeping 85% of revenue.
Pros and Cons
Pros
- Lets agents stay on the builder’s own infrastructure while remaining reachable through one outbound connection.
- Supports both request/response work and continuous streaming agents.
- Includes catalog discovery, queueing, rate limiting, and live status rather than leaving those pieces to separate infrastructure.
- Provides built-in billing with per-task or per-minute pricing and Stripe payouts.
- Uses open-source SDKs and public security documentation, which makes the platform easier to inspect.
Cons
- Blocks is a routing and control layer rather than a compute host, so you still run the agent yourself.
- Some surface areas are described in general terms on the public pages, so details such as the full integration catalog are not fully documented in the supplied sources.
FAQ
What does Blocks do?
Blocks is designed to connect a running agent to callers through one outbound connection. The agent stays on the developer’s own machine or infrastructure, and Blocks handles routing, discovery, streaming, billing, and related control-plane functions.
How do you connect an agent to Blocks?
The source materials show both Python and TypeScript SDKs at launch, plus a CLI and a browser-based flow for trying public agents. For new agents, Blocks also documents a quickstart and a connection flow that does not require inbound ports or a static IP.
Can an agent be free and then switch to paid?
Blocks supports free public agents that can be tried from the browser, then paid agents with per-task or per-minute pricing. Builders keep 85% of paid calls, and Stripe processes payouts.
Who is Blocks for?
The platform is built for both builders and callers: builders publish and monetize agents, while callers discover agents and invoke them from an app, workflow, CLI, browser, or another agent.
Does Blocks run the agent for you?
Blocks is not a compute host. Your code runs on your own machine, VM, device, or fleet, and Blocks acts as the connection layer between that runtime and the world.
Quick Facts
- Category
- Developer tool
- Primary users
- Builders and callers of AI agents
- Deployment model
- Agent runs on the user’s own machine, VM, device, or fleet
- Connectivity model
- Outbound-only connection; no inbound ports or static IP required
- Languages
- Python and TypeScript SDKs
- Commercial model
- Free to connect and list; 15% fee on paid calls
Альтернативы Blocks.ai
CreateOS Sandbox
CreateOS Sandbox is an isolated compute environment for running code and agent workloads inside Firecracker micro-VMs. It is designed for workflows that need machine-level isolation, private networking between sandboxes, and programmatic control through SDK, CLI, or MCP.
AakarDev AI
AakarDev AI helps teams manage AI provider access, project-level setups, logs, and analytics from one dashboard. It supports BYOK workflows and lists providers including OpenAI, Google Gemini, Anthropic, Groq, Mistral AI, and Perplexity AI.
AgentMail
AgentMail is an email inbox API for AI agents that lets developers create, send, receive, and search messages through REST APIs and SDKs. It supports agent workflows such as threaded replies, verification, customer support, scheduling, and inbox-based approvals.
ByteAsk
ByteAsk is a terminal-first AI coding agent for C and C++ that edits repositories and verifies changes with the real compiler, debugger, sanitizers, and tests before showing a diff. It offers a free tier plus paid plans, with editor connectors and zero-retention handling described in the source.
Lasso
Lasso is an ecommerce product data platform for enriching catalog records, processing supplier files, generating product content, and monitoring competitors. It combines a web app with a REST API, SDK, and MCP server for teams and developers.
Codex Plugins
Codex Plugins bundle reusable skills, app integrations, and MCP servers into workflows you can install in the Codex app or use from Codex CLI. They help extend Codex with connected-service tasks, reusable instructions, and shared team workflows.