Quick Answer: Dub Partners is the affiliate program product from Dub, the link infrastructure company. It reached general availability in August 2025 and, by its own published counters as of August 2026, has processed $34M+ in partner commissions across $178M+ in partner-driven revenue, with roughly $2M in payouts moving per month. Pricing is a flat subscription ($90/mo entry) plus a payout processing fee of 3 to 5 percent, with no percentage taken of partner-driven revenue. The API supports per-partner time-series reporting, normalized USD amounts, and working date filters, which avoids several failure modes we have documented in PartnerStack's API. The caveats: the Partners product is commercially licensed rather than open source, the partner network is small (7,000+ active partners against Impact.com's 90,000), there is no multi-motion PRM, and the platform is one year old with a core engineering team of three to five people.

Every few years a platform enters the affiliate software market with enough momentum to change what buyers expect. PartnerStack did it for B2B SaaS marketplaces. Impact.com did it for enterprise partnership breadth. The current candidate is Dub Partners, and unlike most challengers, it leaves enough public evidence (an open repository, published pricing, live counters) that the claims can be checked rather than taken on faith.

This analysis draws on the dubinc/dub repository (commit history, merged pull requests, license files, and the full source tree), Dub's API documentation and help center, and the pricing pages of six competing platforms. Everything was pulled on August 19 and 20, 2026. Where a number comes from a vendor's own counter or a third-party estimate, we say so.

What Dub Partners is

Dub started as link infrastructure: short links, QR codes, and click analytics, used by marketing teams at Framer, Perplexity, Twilio, and Buffer. Dub Partners sits on top of that attribution layer. It ran a nine-month beta that tracked over $9M in affiliate revenue with Framer, Superhuman, Granola, and Tella, then went generally available in August 2025 on the $90/mo Business plan and above.

The product scope as of August 2026:

• Server-side click, lead, and sale tracking (the referral is read from the URL before any JavaScript loads, so ad blockers and Safari cookie expiry do not break attribution)
• Reward structures per click, lead, or sale, plus bounties and email campaigns targeted by partner tags
• Partner groups, applications, messaging, and a discovery network of 7,000+ active partners
• Payouts to up to 100,000 partners from one consolidated invoice, with tax compliance handled (W-9, 1099, W-8)
• White-label referral dashboards that embed in your own product

The traction numbers, with dates

Vendor counters deserve skepticism, so here they are with timestamps rather than adjectives. At general availability in August 2025, Dub reported $9M in affiliate revenue tracked during beta. By December 15, 2025 it reported $10M in cumulative partner payouts. By August 2026 the public counters read $34M+ in partner commissions earned, $178M+ in partner-driven revenue, and $2M+ in monthly payouts. Taken together, cumulative payouts roughly tripled in eight months.

The flagship case study is Framer, which Dub says moved from $267K paid to 312 affiliates in December 2024 to $784K paid to 649 affiliates in December 2025, and now runs about $900K in monthly partner payouts through the platform, citing 40+ staff hours saved per month on invoicing and tax handling. That is a vendor-published case study and should be read as one, but the direction and scale are consistent with the platform counters.

A pricing model no incumbent shares

Dub monetizes a flat subscription plus a payout processing fee. The Business plan is $90/mo with $2,500 in monthly partner payouts included, a 500-partner cap, and a 5 percent fee on payouts. The Advanced plan is $300/mo with $15,000 in monthly payouts included, unlimited partners, and a 3 percent fee. Enterprise is custom. There is no percentage taken of partner-driven revenue at any tier.

That structure is different from everything else in the field:

A worked example makes the difference concrete. Take a program driving $200K per month in partner-attributed revenue at a 25 percent commission, so $50K per month in payouts:

• Dub Advanced: $300 + 3% of $50K = roughly $1,800/mo
• Impact Essentials: $500+ base + 2.5% of $200K = roughly $5,500+/mo
• PartnerStack: subscription plus the reported 3 to 15% of $50K = roughly $2,000 to $9,500/mo (quote-gated, unverified)
• Rewardful Enterprise: about $149/mo flat, but payouts, tax forms, and compliance remain your labor, which is exactly what Dub is selling back

The fee bases differ (payouts versus revenue), so the comparison moves with the commission rate: the higher the revenue share, the better revenue-based pricing looks. At the 20 to 30 percent rates typical of B2B SaaS programs, a flat fee plus a payout percentage costs materially less at scale.

The API avoids the traps we keep finding elsewhere

We recently documented seven PartnerStack API behaviors that break affiliate reporting. The natural question for any new platform is whether it repeats them. We checked Dub's documentation and source against that exact list. Point by point, mostly no.

• Per-partner clicks over time. PartnerStack exposes only a lifetime cumulative click counter per partner. Dub's /analytics endpoint supports groupBy=timeseries filtered by partner, with hourly buckets for short ranges, daily up to 180 days, and monthly beyond, plus a raw /events export with CSV download. This is the single biggest reporting difference.
• Date filtering. Analytics, events, and commissions all take interval presets or custom start and end dates with a timezone parameter.
• Currency. Non-USD sales are converted at ingestion and stored as USD cents everywhere. No mixed-currency sums.
• Pagination. Page-number based on nearly every endpoint (commissions uses a bidirectional cursor). A failed page can simply be re-requested, so there is no equivalent of a bad record stranding everything older than it.
• Rate limits. Published per plan (600 requests/min on Pro up to 3,000/min on Advanced) with explicit 429 responses and Retry-After headers rather than silent failures.
• Webhooks. Eleven event types covering clicks, leads, sales, commissions, and partner lifecycle, with exponential backoff retries and auto-disable after 20 consecutive failures.

The gaps, because there are real ones. Partner records carry lifetime-only rollup totals, so period reporting must come from the analytics and events endpoints (workable, unlike PartnerStack, but still a second query). List responses do not return total counts. There are no webhook events for commission updates or payouts, so refunds and clawbacks require polling. The dashboard has no cohort views at all; if cohort LTV is part of your reporting, you are building it from raw events. Tracked events are capped per plan (250K per month on Business, 1M on Advanced). And twelve months of public GitHub issues show no reports of wrong analytics numbers or broken pagination, which is encouraging but weak evidence, since hosted customers report through support and one-year-old infrastructure has not yet met the traffic that surfaces the ugly edge cases.

What the codebase says about intent

A platform tells you who it intends to take customers from by what it builds importers for. The Dub repository contains dedicated import code for PartnerStack (11 files), Rewardful, Tolt, Tapfiliate, and FirstPromoter, alongside published migration guides for PartnerStack and Rewardful.

The commit history says the company has bet itself on this product. In a sample of 161 pull requests merged between mid-July and mid-August 2026, roughly 70 percent were Partners work: reward tooling, payout rails (Stripe transfers, PayPal, gift cards, a SEPA rollout, stablecoin payouts), fraud controls (cross-program bans, disposable email blocking, identity verification through Veriff), and partner messaging. Partner-related code is now the largest surface in the repository.

Worth weighing against that: nearly all of it comes from three to five engineers. That concentration buys speed and coherence, and it is also a thin team to be carrying a platform that moves other companies' money.

The fine print

Three things a buyer should know before repeating the phrase 'open source affiliate platform.'

First, the license. The repository root is AGPLv3, but the entire Partners product (the partner portal, program management, payouts, bounties, campaigns, fraud tooling, and the lead and sale tracking endpoints) lives in directories governed by the Dub.co Commercial License. Production self-hosting of those directories without a paid agreement is prohibited; development and testing are allowed. The genuinely open source core is link shortening, QR codes, and link analytics. The code being readable is valuable for diligence, and it is not the same thing as being free to run.

Second, the API edges. The partner-side API is marked 'Coming soon' in Dub's own docs, and the payouts API is read-only. Programs that need to create payouts or bounties programmatically cannot yet.

Third, the network. 7,000+ active partners is useful and growing, but recruitment-driven programs lean on marketplace depth, and PartnerStack's B2B marketplace and Impact's 90,000-partner network are still a different weight class.

Where it fits, and where it does not

The fit is clear: SMB and mid-market B2B SaaS, especially Stripe-billed products where the finance team is doing affiliate invoicing by hand and nobody wants to own tax forms. In that seat, Dub competes on price with Rewardful, Tolt, and FirstPromoter while bundling payout operations none of them fully handle, and its API supports the kind of reporting that otherwise requires workarounds. Platform selection is one of the first calls we make in an affiliate program engagement, and as of this year Dub is on that shortlist.

The non-fits are just as clear. Programs built on marketplace recruitment should stay on PartnerStack. Multi-motion programs (resellers, referral partners with deal registration, MDF) need PartnerStack's enterprise tier or Impact. Consumer brands running coupon and media publisher motions belong on the networks. And risk-averse enterprises will reasonably wait for a longer compliance track record before routing seven-figure annual payouts through a one-year-old product.

Frequently Asked Questions

Is Dub Partners open source?

Not in the sense most people mean. The Dub repository is public and the core link infrastructure is AGPLv3, but the Partners product sits in commercially licensed directories, and running that code in production without a paid license is prohibited. It is source-available for reading and diligence, not free to self-host as an affiliate platform.

How much does Dub Partners cost?

As of August 2026: the Business plan is $90/mo with $2,500 in monthly partner payouts included, up to 500 partners, and a 5 percent payout processing fee. The Advanced plan is $300/mo with $15,000 in monthly payouts included, unlimited partners, and a 3 percent fee. Enterprise is custom. Annual prepay carries a 10 percent discount.

Does Dub take a percentage of affiliate revenue?

No. Dub charges the subscription plus a processing fee on payouts. Impact.com charges 2.5 percent of partner-driven transactions, Awin charges a 2.5 to 3.5 percent per-transaction override, and PartnerStack reportedly charges 3 to 15 percent of partner commissions. That structural difference is most of Dub's cost advantage at scale.

Can Dub Partners replace PartnerStack?

It depends on which PartnerStack you use. For tracking, commissions, and payouts on an affiliate motion, yes, and Dub ships a PartnerStack importer and migration guide for exactly that move. For marketplace-driven partner recruitment, reseller programs, or deal registration, no. Migrations also carry their own risks around double payment and attribution continuity regardless of destination platform, so plan them as projects, not toggles.

Does the Dub API support per-partner click reporting by date range?

Yes. The analytics endpoint returns time-series data filterable by partner, with hourly granularity up to two days, daily up to 180 days, and monthly beyond, and a raw events endpoint exports individual click, lead, and sale events. This is the capability whose absence in PartnerStack forces daily snapshot-diffing workarounds.

Is a one-year-old platform safe to run payouts through?

That is a judgment call, and the evidence points both ways. Dub processes over $2M in monthly payouts, runs identity verification and fraud controls, and publishes retry and failure semantics for its integrations. It is also five engineers' work with a short compliance track record. A reasonable middle path is piloting with one partner segment, keeping raw event exports flowing to your own storage, and holding the incumbent program open until a full payout cycle has reconciled cleanly.

Choosing between Dub, PartnerStack, Rewardful, and Impact is a decision about payout volume, partner mix, and sales motion, not feature lists. Contact Jolly Consulting for a platform assessment grounded in your program's actual economics.

Related reading:

• 7 PartnerStack API Behaviors That Break Affiliate Reporting: The failure modes this post benchmarks Dub against
• Impact vs PartnerStack for B2B: The incumbent comparison at the mid-market and enterprise end
• The 7 Best Affiliate Platforms for B2B SaaS: Platform selection ranked by company stage