Pipedream
Pipedream is the integration layer for AI agents: managed auth and 10,000+ tools across 3,000+ APIs.
For any agent that has to touch a user's own accounts, Pipedream's Connect removes the OAuth, token-vault, and refresh work that would eat your roadmap for a quarter. The remote MCP server, OAuth on a static URL, and a real raw proxy mean you can go prebuilt-first and drop to a raw request when nothing fits. The free tier runs everything in development with no credit card, but production needs Startup at $150/mo billed monthly or $99/mo billed annually, plus $2 per external user past 100. If nobody on your team ships code, this is the wrong shelf — look at Zapier or Make.
Verified 2d ago · liveness 95/100 · cite: rightaichoice.com/tools/pipedream
- Developers building AI agents that act inside each user's own Slack, Gmail, Notion, or Stripe accounts
- SaaS teams that want an integration catalog in days instead of registering OAuth apps
- Platform teams that want MCP tool discovery plus a raw proxy for uncovered endpoints
- Teams that need credentials encrypted at rest and scoped per project for security review
- Non-technical users who want drag-and-drop automation with no code
- Teams whose core need is dedicated low-latency streaming rather than webhook triggers
- Projects that only ever call unauthenticated public APIs — the auth layer earns nothing there
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Pipedream if you want a no-code builder where business users wire up automations by clicking, or if your integration surface is a handful of unauthenticated API calls where managed OAuth buys you nothing.
The free tier is development only — the moment you serve real external users you need Startup at $150/mo billed monthly, or $99/mo billed annually, which is a step up from $0.
Free covers the whole toolkit in development with no credit card, which is genuinely usable for a solo developer prototyping an agent. Startup at $150/mo billed monthly (or $99/mo billed annually) is priced for a funded team shipping to real users, and it undercuts hiring even one engineer to build OAuth. Past 100 external users the $2-per-user rate and $0.012-per-credit overage make usage math matter; at that scale compare against building in-house or a workflow vendor's enterprise plan.
In short
Pipedream — Pipedream is the integration layer for AI agents: managed auth and 10,000+ tools across 3,000+ APIs. Best for Developers building AI agents that act inside each user's own Slack, Gmail, Notion, or Stripe accounts, SaaS teams that want an integration catalog in days instead of registering OAuth apps, Platform teams that want MCP tool discovery plus a raw proxy for uncovered endpoints. Free to start; paid plans from $150/mo.
What's new in Pipedream
Checked 2 days agoAcross the latest 5 updates: 4 feature updates and 1 launch.
OAuth for Pipedream MCP + ChatGPT Support
MCP server now uses OAuth on a static URL at mcp.pipedream.net/v2 and works inside ChatGPT. All 10,000+ tools gained toolAnnotations marking read vs write and destructive operations.
Python, TypeScript, and Java SDKs + Interactive AI-Friendly Docs
First Python and Java SDKs for Connect shipped, TypeScript SDK rewritten as v2 with namespaced methods. New interactive API reference and AI-friendly docs published; v1 remains supported.
Edit Existing Workflows with AI
Edit with AI button in the workflow builder lets you change whole workflows or single code steps with natural language, with noted limits on control flow, GitHub sync, and Python steps.
MCP Chat - Talk directly with all your tools
Open-sourced MIT MCP Chat app (chat.pipedream.com) powered by Pipedream Connect and Vercel's AI SDK, giving direct access to 10k+ tools from 2,800+ APIs with auth built in.
Remote MCP Server and OpenAI Support
Pipedream shipped a remote MCP server for developers exposing tools to AI apps, along with support for OpenAI clients.
What people actually say about Pipedream — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
7 mentions across 5 sources (Hacker News, YouTube, Product Hunt, GitHub, Lemmy), 58 more we could not attribute · researched Oct 10, 2026.
Weighted by the 65 posts each of 5 sources contributed.
- +Managed OAuth for 3,000+ APIs means you never touch customer credentials yourself
- +10,000+ prebuilt tools from a source-available GitHub registry accelerates agent development
- +Remote MCP server at mcp.pipedream.net/v2 works in ChatGPT and other MCP clients
- +toolAnnotations on all tools flag read vs write vs destructive actions — useful for agent safety
- +SDKs in TypeScript, Python and Java keep it provider- and framework-agnostic
- −Explicitly not drag-and-drop — requires SDKs and code, unlike Zapier or Make
- −Public YouTube complaint about Pipedream 'ghosting customers' on support
- −4,515 open GitHub issues against 11,720 stars signals a heavy unaddressed backlog
- −Community-submitted app requests (ShapelessAI, FractalAI, NexaVoxa) sit open and unanswered
- −Conduit enterprise gateway — SSO, audit trail — still in early access, not GA
- • Usage-based invocation pricing can scale unpredictably with agent traffic
- • Conduit enterprise features are gated behind early access — no public price
- • Support responsiveness concerns suggest you may need internal capacity to compensate
Viability Score
How well maintained and how widely used is Pipedream? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: October 2026
How we score →Key Features
- Managed auth for 3,000+ APIs with hosted OAuth clients and automatic token refresh
- 10,000+ prebuilt tools and triggers from a source-available registry on GitHub
- Remote MCP server at mcp.pipedream.net/v2 with OAuth on a static URL
- MCP server works inside ChatGPT and Claude as well as any MCP client
- toolAnnotations on all 10,000+ tools marking read vs write and destructive operations
- Connect Link hosted authorization flow you can drop in instead of building UI
- Browser SDK and connect-react package to embed the connect flow in your app
- Raw proxy that attaches a user's credentials to any endpoint and HTTP method
- Real-time triggers that webhook events from your users' apps to your URL
- Deploy triggers per external user with a single SDK call
- TypeScript SDK v2 (rewritten, namespaced methods), Python SDK, and Java SDK
- REST API and CLI
- Bring your own OAuth clients when you want them
- Credentials encrypted at rest, scoped per project, never exposed to your app
- Revoke a connected account with one API call
About Pipedream
Pipedream gives your AI agent tools and credentials. Point any MCP client at Pipedream's remote server at mcp.pipedream.net/v2 and the agent gets tool discovery plus per-user auth in a single connection; it works in ChatGPT and Claude as well as in your own agent code. Underneath sits Connect, the developer product that runs the whole authorization lifecycle: hosted OAuth clients, hosted connect flows through Connect Link, encrypted token storage, and automatic refresh, so you never handle a user credential yourself. On top of that auth layer sit 3,000+ connectors exposed as 10,000+ prebuilt tools and triggers from a source-available registry on GitHub, plus a raw proxy that attaches a user's credentials to any endpoint and HTTP method on those APIs. Language coverage is TypeScript (SDK v2, rewritten with namespaced methods), Python, and Java, alongside REST and a CLI. For organizations, Conduit is a gateway in early access with SSO, access policies, per-user permissions, and an audit trail. This is SDK-and-code infrastructure, not drag-and-drop automation: it suits developers embedding a connector catalog into their own product or agent, and it does not suit no-code teams who want Zapier or Make.
Behind the Verdict
Pipedream's pitch has narrowed and sharpened: it is now the integration layer for AI agents rather than a general automation canvas. That focus shows up in the product. The remote MCP server is the front door — one URL, per-user auth via the x-pd-external-user-id header, tool discovery built in, and as of October 1, 2025 OAuth authentication on a static URL plus support inside ChatGPT. All 10,000+ tools carry toolAnnotations marking read vs write and destructive operations, which matters when your agent has to choose a tool unattended. Behind the MCP layer, Connect handles the parts that are genuinely painful to build. Hosted OAuth clients mean you don't register your own apps for 3,000+ APIs. Connect Link gives you a hosted authorization page you can drop in instead of building connect UI; the browser SDK and connect-react package let you embed the flow if you'd rather own the surface. Credentials are encrypted at rest, scoped per project, never exposed to your app, and revocable with one API call. Fresh tokens are injected per user on every request in Pipedream's managed runtime. The tools and triggers layer is where the breadth lives. Actions write to APIs with per-user auth in a single call; triggers deploy from the registry and webhook events from a user's apps to your URL in real time — the docs show a Gmail new-email trigger deployed with one SDK call. Anything not covered is reachable through the Connect proxy, which takes any endpoint and HTTP method on any of the 3,000+ APIs and attaches the right credentials server-side. The SDK story matured on September 9, 2025 with first-time Python and Java SDKs and a rewritten TypeScript SDK v2 with namespaced methods and better type definitions; v1 remains supported, so existing implementations aren't stranded. Documentation was rebuilt as an interactive API reference with snippets for REST, TypeScript, Python, Java, and Go, plus an AI-native path — docs as plain text or markdown, or a docs MCP server for your IDE or coding agent. Where it strains: this is infrastructure, and it reads like infrastructure. The value arrives through SDKs, the CLI, the REST API, or MCP, not a no-code canvas. Conduit — the enterprise gateway with SSO, SCIM, HIPAA, static egress IPs, and audit trails — is still labeled early access on the homepage and pricing page, so governance-heavy buyers should plan around that status rather than assume it. The AI workflow editor shipped September 2, 2025 works for most workflows but the changelog itself flags limitations with control flow, GitHub sync, Python steps, and Pipedream Connect features. Triggers deliver events to a webhook; if you need dedicated low-latency streaming, that's a different product. And the platform is deliberately framework and LLM-provider agnostic, so it bundles no model — you bring your own agent stack. The honest comparison: Zapier and Make win on no-code reach and prebuilt app breadth for business users; Pipedream wins when your integration
Researching Pipedream? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Pipedream actually fits — and what changes day-one when you adopt it.
Point your MCP client at mcp.pipedream.net/v2 with the user's email in x-pd-external-user-id, then call slack-send-message and gmail-new-email-received without registering a single OAuth app.
Outcome: A working per-user agent in development on the free tier with no credit card, before you commit to any paid plan.
Use pd.tokens.create to mint a Connect Link, send the user through the hosted OAuth flow, then call pd.proxy.post on a Notion endpoint the registry doesn't cover.
Outcome: A connector catalog shipped in days instead of a quarter of OAuth work, with tokens encrypted at rest and revocable by API call.
Evaluate Conduit's SSO, access policies, per-user permissions, and audit trail against your governance requirements during its early-access period.
Outcome: A decision on whether to deploy now on early-access terms or wait for General Availability, made with the early-access labeling in view.
Use Cases
- Give an AI agent 10,000+ tools through one MCP endpoint with per-user authentication
- Embed a connector catalog into your SaaS product without registering OAuth apps
- Invoice the week's new Stripe customers and post a summary to a Slack channel from an agent
- Deploy a Gmail trigger that webhooks new-email events to your app in real time
- Send any request to an endpoint with no prebuilt tool through the Connect proxy with auth injected
- Let employees connect their work apps to AI agents under centrally managed policies with Conduit
- Summarize recent Linear tickets or draft a release announcement through MCP Chat
- Inspect Connect SDK calls with payloads and responses in the developer playground
Models Under the Hood
as of 2026-10-16
Limitations
- Pipedream is developer-focused: the value arrives through SDKs, the CLI, the REST API, or MCP, not a no-code canvas, so read the Connect quickstart before assuming a weekend plug-in.
- AI workflow editing, shipped September 2, 2025, covers most workflows but the changelog flags limitations with control flow, GitHub sync, Python steps, and Pipedream Connect features.
- Triggers deliver events to a webhook rather than providing dedicated low-latency streaming infrastructure.
- Conduit, the enterprise gateway with SSO, access policies, per-user permissions, and audit trails, is still marked early access on the homepage, so governance-heavy buyers should plan around that status.
- The platform is framework and LLM-provider agnostic by design, so it does not bundle a model — you bring your own agent stack.
as of 2026-10-19
Verification history
We have re-verified Pipedream 20 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-checked, vendor evidence unchanged
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 20 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Pipedream tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Free
$0
Ideal for
Solo developer or small team prototyping an agent or integration catalog before any real users touch it
What this tier adds
Free entry point: every feature and all 3,000+ connectors, but development environment only, no credit card required
Startup
$150/mo billed monthly, or $99/mo billed annually
Ideal for
Funded team shipping an agent or SaaS integration catalog to real external users
What this tier adds
Adds the production environment, 100 external users included (then $2 each) and 10,000 monthly usage credits (then $0.012 each)
Business
Custom
Ideal for
Enterprise or regulated company that needs SSO, HIPAA, static egress IPs, and contractual SLAs
What this tier adds
Adds volume pricing, SSO, SCIM, HIPAA, dedicated static egress IPs, SLAs, and dedicated Slack support on an annual contract
Where the pricing makes sense
The company stage and team size where Pipedream's pricing actually pencils out — and where peers do it cheaper.
Free covers the whole toolkit in development with no credit card, which is genuinely usable for a solo developer prototyping an agent. Startup at $150/mo billed monthly (or $99/mo billed annually) is priced for a funded team shipping to real users, and it undercuts hiring even one engineer to build OAuth. Past 100 external users the $2-per-user rate and $0.012-per-credit overage make usage math matter; at that scale compare against building in-house or a workflow vendor's enterprise plan.
Setup time & first value
How long it actually takes to get something useful out of Pipedream — broken out by persona, not the marketing-page minute.
A solo developer can go from npm i @pipedream/sdk to a first per-user MCP call in an afternoon — the OAuth dance is handled for you. A product team embedding Connect Link or connect-react into their app should budget a few days including security review. Conduit evaluation runs longer because it is early access and enterprise-gated.
Switching to or from Pipedream
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From self-built OAuth: swap your token store for pd.tokens.create and the Connect proxy, keeping your existing app UI.
- →From Zapier or Make: rewrite Zaps/scenarios as Connect actions and deployed triggers, or call covered endpoints through the Connect proxy.
- →From TypeScript SDK 1.x: follow the published migration guide to SDK v2 with namespaced methods; v1 stays supported so the move isn't forced.
- ↗To Zapier or Make: rebuild automations on their no-code canvas and their own OAuth app registrations.
- ↗To a self-hosted connector stack: export logic and re-register every OAuth app yourself, replacing Pipedream's hosted credentials and refresh.
Integrations
Resources & Guides
- Documentationpipedream.com
Introduction to Pipedream
Pipedream provides the toolkit to add thousands of integrations to your app and enables you to automate any process.
- Quickstartpipedream.com
Workflow Development
Pipedream provides the toolkit to add thousands of integrations to your app and enables you to automate any process.
- Quickstartpipedream.com
Connect Quickstart
Pipedream provides the toolkit to add thousands of integrations to your app and enables you to automate any process.
- Documentationpipedream.com
Introduction
Pipedream provides the toolkit to add thousands of integrations to your app and enables you to automate any process.
- Documentationpipedream.com
Llms
Full product docs from pipedream.com
- Resourcepipedream.com
One SDK, thousands of API integrations
Pipedream Connect provides managed authentication for 3,000+ APIs and access to 10,000+ prebuilt tools. Roll your own frontend with the server SDK or use @pipedream/connect-react to get started in minutes.
- Resourcepipedream.com
Blog
Connect APIs, remarkably fast.
- Resourcepipedream.com
Pipedream
Connect APIs, remarkably fast.
Tutorials & Learning

Pipedream | 正直なレビュー(知っておくべきことすべて)
How To Mr

Pipedream Review 2026 – Powerful Serverless Integration & Workflow Automation Platform
Bytes Media
YouTube returned 6 videos for “Pipedream”, and we withheld 4: 4 could not be judged, because “Pipedream” is a single word that other videos use for other things. Showing the 2 we can prove are about Pipedream.
Official links
Tools that pair well with Pipedream
Common stack mates teams adopt alongside Pipedream, with the specific reason each pairing earns its keep.
Zhipu GLM
Zhipu GLM is a full-stack large-model platform built on open GLM weights, with MaaS APIs, ZCode and shipped computer-use agents.
Context7
Context7 feeds version-pinned library docs into Claude, Cursor, and Codex agents so they stop inventing APIs.
Chrome DevTools MCP
Apache-2.0 MCP server that gives coding agents a live Chrome DevTools session for debugging, traces, and browser automation.
Featured Head-to-Head Comparisons
Alternatives to Pipedream
View allZhipu GLM
Zhipu GLM is a full-stack large-model platform built on open GLM weights, with MaaS APIs, ZCode and shipped computer-use agents.
Context7
Context7 feeds version-pinned library docs into Claude, Cursor, and Codex agents so they stop inventing APIs.
Chrome DevTools MCP
Apache-2.0 MCP server that gives coding agents a live Chrome DevTools session for debugging, traces, and browser automation.
Frequently Asked Questions
Used Pipedream? Help shape our editorial sentiment research.