Pipedream pricing

About

Pipedream is a developer-first workflow-automation and integration platform. It lets engineers build event-driven workflows that connect 3,000+ integrated apps, run arbitrary Node.js, Python, Go, or Bash code between steps, and expose those workflows as APIs, schedules, or webhook handlers. Its self-serve, code-friendly positioning places it between low-code automation tools (Zapier, Make) and raw serverless platforms — it markets itself to the million-plus developers who want programmability without managing infrastructure.

The product has two faces. The core is workflow automation: hosted, event-driven pipelines billed on compute. The second is Pipedream Connect, a managed-authentication and tool-execution layer that developers embed in their own applications and AI agents — handling OAuth for end users and exposing 10,000+ pre-built tools, an API proxy, and deployable MCP servers. Connect is sold as its own plan tier, reflecting a deliberate move to monetize Pipedream as embeddable infrastructure for agentic products, not just a standalone automation console.

Pipedream operates out of San Francisco and reports being trusted by 1,000,000+ developers from startups to Fortune 500 companies. The company was acquired by Workday (the “Pipedream has joined Workday” banner sits atop the live pricing page as of June 2026), folding a usage-priced, developer-led automation and embeddable-integration platform into a large enterprise HR and finance suite. Notably, the acquisition has not (yet) collapsed Pipedream’s self-serve pricing: the public Free → Business tiers, the credit meter, and the embeddable Connect product all remain live and unchanged. For the live pricing surface this blueprint analyzes, see Pipedream’s pricing page.

Pricing summary : credit-metered compute under a tiered subscription

Pipedream runs a hybrid pricing model: a tiered recurring subscription on top of credit-based billing for compute. Each plan bundles a monthly credit allowance, and credits deplete as workflows run — one credit covers 30 seconds of execution at the default 256MB of memory, so cost tracks duration and memory, not step count. A separate, bundled AI-token allowance (1M to 50M tokens/mo) covers built-in AI actions, layering a second token-based meter on top of the credit meter. The free tier converts via product-led, self-serve sign-up; the top Business tier is sales-led.

The model has these dimensions:

• Plan fee — five tiers: Free ($0), Basic ($29/mo annual, $45/mo monthly), Advanced ($49/mo annual, $74/mo monthly), Connect ($99/mo annual, $150/mo monthly), and Business (custom).
• Included credits/mo — 100 (Free, hard-capped) · 2,000 (Basic) · 2,000 (Advanced) · 10,000 (Connect) · Custom (Business). Paid plans remove the usage cap; the Free plan stops at 100.
• Included AI tokens/mo — 1M (Free) · 20M (Basic) · 50M (Advanced) · 50M (Connect) · Custom (Business), for built-in AI steps.
• Embedded-integration seats — the Connect tier authenticates 100 external end-users and charges $2 per additional external user.
• Capacity limits — active workflows (3 → unlimited), connected accounts (3 → unlimited), max memory (up to 10GB), and max execution duration (300s on Free, 750s on paid) all step up by tier.

What makes this different: the meter is compute time × memory expressed as credits, decoupled from both invocation count and workflow complexity — a developer can chain 50 steps and pay the same as a single long-running step, because only elapsed compute matters. Pipedream also splits its embeddable “Connect” product into its own paid tier, monetizing the auth-and-tools layer that powers other companies’ AI agents.

Pricing by product

Workflow automation (Individual plans)

Workflow automation (Team plans)

Pipedream Connect (embeddable integration product)

Sales motions across products: PLG / self-serve for Free, Basic, Advanced, and Connect (sign up and pay online); sales-led for Business (custom quote, contract, and invoicing).

Hidden costs : What Pipedream users actually pay

The sticker price on a Pipedream plan is rarely the full bill, because the plan fee buys a credit allowance, not unlimited compute. Three things quietly move your real cost above the headline number: metered credit overages once you exhaust the included pool, AI-token consumption beyond the bundled allowance, and per-external-user charges on the Connect tier.

The trap most teams hit is that credits are billed on compute time × memory, not per run. A workflow that polls an API every minute and runs for 20 seconds at 256MB looks cheap, but bump the memory to 2GB to handle a large payload and the same 20 seconds now burns ~8× the credits. Because development and testing are free, the credit burn only shows up once a workflow is live and firing on real volume — so the bill arrives a step removed from the moment you built the thing.

Here is what a representative team actually pays on the Advanced plan ($49/mo annual, 2,000 included credits, 50M AI tokens):

A few cost notes:

• No overage rate on the rate card. The pricing page advertises included credits and “no usage limit” on paid tiers but does not publish a per-credit overage price — additional credits and additional AI tokens are metered (“Use additional AI tokens” is a feature row), so true overage cost is visible only inside the usage dashboard, not the public page.
• Connect adds per-user cost. The Connect tier ($99/mo annual) bundles managed auth for 100 external end-users; beyond that it is $2 per additional external user — a real per-seat-style line item that scales with your customers, not your team.
• Annual lock-in is the cheaper path. Switching from monthly to annual saves $192 (Basic), $300 (Advanced), and $612 (Connect) per year — the monthly convenience premium is roughly 50%.

Want to estimate your own Pipedream bill? Use the Pipedream pricing calculator to model your costs based on usage patterns.

Pricing evolution : Pipedream pricing history and changes

Pipedream’s pricing has evolved through one defining structural shift: a move away from pure consumption pricing (early Pipedream metered raw “invocations” and compute seconds) toward the current credit + tier model that normalizes compute time × memory into a single credit unit. The other landmark is corporate, not commercial — the Workday acquisition — which has so far left the rate card untouched.

Cadence

Tracked range: 2024–present. Earlier per-invocation pricing predates the current credit model.

Notable changes

• ~2024–2025 — Shift from raw consumption (invocations / compute-seconds) to the credit meter, where one credit covers 30 seconds of execution at the default 256MB and scales with memory. This made cost predictable per-unit while keeping it usage-based.
• 2025 — Pipedream Connect carved out as a distinct paid tier ($99/mo annual), monetizing the managed-authentication and 10,000-tool layer that powers embedded integrations and AI agents — a deliberate productization of infrastructure other companies build on.
• 2025 — Acquired by Workday. The standalone developer pricing (Free → Business), the credit meter, and Connect all remained live and unchanged through the transition.
• 2026-06-16 — Live rate card verified: Free $0 (100 credits, hard cap), Basic $29/mo, Advanced $49/mo, Connect $99/mo (all annual), Business custom; month-to-month $45/$74/$150.

What’s unique : Pipedream’s distinctive pricing mechanics

1. The meter is compute time × memory, not invocations or steps. Most automation tools charge per task, per run, or per step — Zapier counts “tasks,” Make counts “operations.” Pipedream instead bills a single credit = 30 seconds at 256MB, and memory scales it linearly. A developer can chain 50 steps into one workflow and pay the same as a single long step, because only elapsed compute matters. This rewards efficient code and punishes long-running or memory-hungry workflows — a pricing shape that looks far more like a serverless platform (Lambda’s GB-seconds) than like a no-code automation tool.

2. Connect is a product sold as a pricing tier. Pipedream took the managed-authentication and tool-execution layer that powers embedded integrations — OAuth for your end users, 10,000+ pre-built tools, an API proxy, deployable MCP servers — and made it its own $99/mo plan with a per-external-user meter ($2 beyond the first 100). Few competitors price the embeddable-infrastructure use case as a distinct SKU; most bury it in enterprise. This is a bet that AI-agent builders will pay for auth-and-tools as a primitive.

3. Bundled AI tokens as a second, separate meter. On top of the credit meter, every tier bundles an AI-token allowance (1M → 50M/mo) for built-in AI steps, with additional tokens metered separately. Pipedream is effectively reselling model inference inside the workflow product, layering a token meter on a credit meter — two usage dimensions in one bill, each with its own included allowance.

Strengths & weaknesses

Billing UX : credit allowances, usage dashboards, and invoicing controls

• Annual / Monthly billing toggle — the pricing page exposes a switch between annual and month-to-month billing; annual is the default and is materially cheaper (Basic $29/mo vs $45/mo, Advanced $49/mo vs $74/mo, Connect $99/mo vs $150/mo).
• Credit allowance + included-credits slider — each plan card shows the included monthly credits (100 → 10,000) on a 0-to-10M+ usage slider, signaling that paid plans meter additional usage above the included pool rather than hard-stopping.
• Free-tier hard cap — the Free plan enforces a 100-credit/month usage cap (no overage), an explicit guardrail against runaway spend for non-paying accounts.
• Usage breakdown view — Pipedream’s FAQ confirms a per-account usage breakdown (“Where can I see a breakdown of my usage?”) so customers can attribute credit and token consumption.
• Invoicing and payment options — paid plans bill by credit/debit card by default; the Business tier adds pay-by-invoice, pay-by-bank-transfer, custom invoicing, and the ability to edit billing email, VAT, and company details tied to invoices.
• Event History retention — execution/event history is retained 7 days (Free/Basic), 30 days (Advanced/Connect), and custom (Business), governing how far back users can inspect billed activity.

Strategic wins : Why Pipedream’s pricing decisions worked

1. Picking compute, not tasks, as the value metric

By metering on compute time × memory rather than per-task, Pipedream aligned its pricing with what it actually pays for (hosted execution) and with how developers think about cost (serverless GB-seconds). This neutralizes the biggest objection to no-code automation tools — that complex multi-step workflows get expensive fast — and lets power users build elaborate pipelines without watching a task counter. It is a textbook case of choosing a value metric that scales with the customer’s value, not their headcount. See usage-based pricing strategy for the framework.

2. A free tier built to convert, not just to demo

The Free plan is unusually complete: 100 credits, 1M AI tokens, 3 live workflows, and unlimited testing, with no credit card. That is enough to ship a real automation, which makes the upgrade to Basic ($29) a natural step once you outgrow the credit cap or need a fourth workflow — classic product-led growth sequencing. The hard 100-credit cap on Free (the only tier with a cap) is the deliberate friction that triggers the upgrade.

3. Productizing the infrastructure layer as Connect

Carving managed auth + 10,000 tools into a standalone $99/mo Connect tier turns a feature into a revenue line and positions Pipedream as embeddable infrastructure for the wave of AI-agent products — exactly the outcome of agent-driven demand other tooling vendors are chasing. The $2-per-external-user meter lets that revenue scale with the customer’s own user base.

Areas to improve : Gaps in Pipedream’s pricing approach

1. Publish the credit overage rate

The page advertises included credits and “no usage limit” on paid tiers, but the actual per-credit (and per-extra-AI-token) overage price is not on the rate card — you only see it once you’re inside the dashboard and consuming. For a usage-priced product, leaving the marginal rate off the public page is the classic setup for bill shock and cost unpredictability. A simple “additional credits: $X / 1,000” line would let buyers self-model.

2. Give buyers a credit-to-dollars estimator

Because a credit is an abstract compute-time × memory unit, most prospects cannot translate “2,000 credits” into “how many workflow runs will that cover for me.” A worked example or an on-page estimator (“a 10-second, 256MB workflow run = ~1/3 credit”) would close the gap between the abstract meter and a real-world bill — the same predictability problem covered in choosing the right usage metric.

3. Clarify the post-acquisition pricing roadmap

With Workday now the owner, self-serve customers have no public signal on whether the Free tier, the credit meter, or month-to-month billing will persist, or whether pricing will migrate toward enterprise contracts. A short commitment statement would reduce the churn risk that uncertainty creates for the developer base that drove Pipedream’s adoption.

Key takeaways

1. Pipedream prices compute, not tasks. The core meter is a credit = 30 seconds of execution at 256MB, scaling with memory — so cost tracks duration and RAM, never step count or number of runs alone. This is serverless-style pricing wearing an automation-tool UI.
2. The tiers are an allowance wrapper around the meter. Free $0 (100 credits, hard cap), Basic $29/mo (2,000), Advanced $49/mo (2,000 + unlimited workflows), Connect $99/mo (10,000), Business custom — each plan fee buys an included credit pool plus a bundled AI-token allowance, with metered overage above it.
3. Connect monetizes embeddable infrastructure. Auth for end users + 10,000 tools + MCP servers are sold as a $99/mo tier with a $2-per-external-user meter — a clean example of turning a platform capability into its own priced SKU for AI-agent builders.
4. Annual billing is the real price. Month-to-month carries a ~33–50% premium ($45/$74/$150 vs $29/$49/$99); Pipedream surfaces the savings as concrete dollar figures to steer buyers to commit.
5. A complete free tier is the growth engine. The Free plan ships real workflows with no card; the 100-credit hard cap is the only deliberate friction, engineered to convert active users to Basic.

UBP implications

1. Normalize your raw cost driver into one legible unit. Pipedream had at least three cost dimensions (time, memory, invocations) and collapsed them into a single “credit.” If your usage meter exposes multiple raw drivers, packaging them into one synthetic unit makes pricing comprehensible — but only if you also publish the conversion so buyers can self-model.
2. An allowance-plus-overage hybrid de-risks usage pricing for buyers. Bundling a generous included pool into each tier gives customers a predictable floor while preserving usage upside. The lesson for UBP teams: the included allowance is your churn insurance; the metered overage is your expansion revenue.
3. Productize the platform layer your customers already lean on. Connect shows that an embeddable capability (auth, tools, MCP) can become a separately priced tier rather than an enterprise feature buried in a contract — capturing the AI-agent buyer with a self-serve price point and a per-end-user meter that scales with their success.

Sources

• Pipedream pricing page (accessed 2026-06-16) — primary rate-card evidence (annual + monthly billing states)
• Pipedream pricing docs — credit definition (accessed 2026-06-16) — “one credit per 30 seconds of compute at 256MB (the default) per workflow segment,” memory-scaled
• Pipedream Connect docs (accessed 2026-06-16) — managed auth, 10k tools, MCP servers, per-external-user pricing
• Pipedream official website (accessed 2026-06-16) — “Pipedream has joined Workday” banner

Bottom line

Pipedream prices like a serverless platform dressed as an automation tool: a tiered subscription (Free $0 → Business custom) wrapped around a credit meter where one credit = 30 seconds of compute at 256MB, scaling with memory rather than step count. The Free tier is complete enough to ship real work, paid entry is cheap ($29–$49/mo), and the Connect tier turns embeddable auth-and-tools into its own $99/mo SKU for AI-agent builders. The model’s strengths — predictable per-unit cost, complexity-agnostic billing, a conversion-grade free tier — are partly offset by an unpublished overage rate and a credit unit that’s abstract for non-developers. The Workday acquisition leaves the rate card intact for now but adds long-term uncertainty for the self-serve base.

Want to compare Pipedream against other developer-tools and automation companies? Browse the pricing blueprint.

Pricing timeline : Major events on a vertical axis

Each milestone below corresponds to a public pricing change, product launch, or material adjustment. Major events use a filled marker; minor adjustments use a faded one.

Facts captured — credit-metered hybrid model, five tiers

Jun 2026

Captured the live pricing page: Free $0, Basic $29/mo, Advanced $49/mo, Connect $99/mo on annual billing and Business custom, with credit-metered compute (1 credit = 30 seconds at 256MB, memory-scaled) plus bundled AI-token allowances. Month-to-month rates are $45, $74, and $150.

Pipedream joins Workday

Sep 2025

Pipedream is acquired by Workday, folding a developer-led, credit-metered automation and embeddable-integration platform into Workday's enterprise HR and finance suite. The standalone self-serve pricing page and tiers remain live and unchanged at acquisition.