Flexprice

Usage-based metering and billing, honestly reviewed. Built for developers who’ve hit the ceiling on percent-of-revenue pricing.

TL;DR

• What it is: Open-source (AGPL-3.0) monetization infrastructure — real-time usage metering, credit management, and flexible billing for AI-native and SaaS teams who need more than Stripe’s flat percentage model [GitHub README][3].
• Who it’s for: Developer teams building AI products or SaaS with usage-based, credit-based, or hybrid pricing. Also teams managing multiple subsidiaries or complex billing structures [5].
• Cost savings: Stripe Billing charges 0.5–0.8% of revenue processed. At $2M ARR that’s $10,000–$16,000/year handed to Stripe for billing logic you can own. Flexprice self-hosted on a decent VPS costs the infrastructure, not a revenue cut [5].
• Key strength: Real-time event metering backed by ClickHouse, with a no-code UI for pricing changes that don’t require engineering tickets. One customer’s Head of Engineering reported billing dropped from 20–30% of a developer’s daily bandwidth to zero [website].
• Key weakness: AGPL-3.0 license means commercial embedding requires care. Young project (about one year old, 3,549 stars). Documentation sources dominate reviews — this is not yet a battle-tested tool with years of independent user reports [3][GitHub].

---

What is Flexprice

Flexprice is a self-hostable billing and metering platform. You instrument your application to send usage events — API calls, tokens consumed, seats provisioned, credits spent — and Flexprice handles everything downstream: aggregating those events in real time, applying pricing rules, managing credit balances, enforcing feature gates, and generating invoices. Its GitHub description is unusually honest: “Usage-based pricing and billing for developers — Cloud or self-hosted, No-code UI, Realtime usage metering, Credits & top-ups, Control feature access” [GitHub README].

What separates it from simple Stripe integrations is the architecture underneath. Flexprice uses ClickHouse as the event store, which means it’s designed for high-volume event pipelines, not the occasional webhook [GitHub README]. PostgreSQL handles relational billing data. The result is a system that can ingest millions of usage events without the counting, aggregating, and deduplication logic living inside your application.

The company’s pitch is explicitly aimed at the pain of building billing in-house: “When existing billing tools don’t flex to your product’s needs, developers shoulder the burden eating up valuable development time and causing ongoing maintenance headaches” [GitHub README]. The alternatives they cite by name on the website are Chargebee, Lago, and Stripe — which tells you the price point and complexity they’re targeting [website].

As of this review: 3,549 stars, 152 forks, AGPL-3.0 license, active development (last commit within 24 hours) [3][GitHub]. The project is about one year old.

---

Why people choose it

The problem Flexprice addresses is specific and real: most billing tools were built for subscriptions, not usage. Stripe Billing, the de-facto baseline, works fine if you charge a flat monthly fee. Once you add usage tiers, credit packs, prepaid balances, per-seat-plus-usage hybrids, or AI token metering, you’re stitching together webhook handlers, Stripe meters, custom database tables, and proration logic. That patchwork is what Flexprice is trying to replace.

Versus Stripe Billing. The clearest case. Stripe Billing charges a percentage of revenue — 0.5% to 0.8% depending on tier [5]. That’s invisible at $50K ARR. At $2M ARR it’s $10,000–$16,000/year for billing plumbing you could own. Stripe’s meters are also limited: custom aggregation logic, credit systems, and real-time balance enforcement aren’t native to Stripe’s billing primitives [5][3]. OpenAlternative lists Flexprice as a direct alternative to Stripe Billing [3].

Versus Orb. Orb is the other commonly cited usage-based billing platform. It’s cloud-only and expensive — positioned for scale, not for teams trying to control infrastructure costs. If you need Orb-level functionality but want self-hosted control and no revenue-share pricing, Flexprice is the logical comparison [3].

Versus Lago. This is the interesting fight because Lago is also open-source, also developer-focused, and has a much larger footprint: 9,541 stars to Flexprice’s 3,549 [3][4]. Lago has been around longer and has broader community validation. Flexprice differentiates on real-time ClickHouse-backed metering, the no-code pricing UI, and credit system depth. Whether those trade-offs are worth choosing the newer, smaller project depends on your team’s risk tolerance.

Versus in-house billing. The most common alternative for engineering teams is building it themselves. The website quotes a Head of Engineering at an unnamed company: “Before Flexprice, billing required one developer to spend 20-30% of their daily bandwidth. After adopting Flexprice, no developer is dedicated to billing at all” [website]. That is the actual pitch — not “it’s cheaper than Stripe” but “stop rebuilding this yourself every time your pricing model changes.”

The multi-entity case. For companies with multiple legal entities — a US parent, EU subsidiary, regional divisions — billing becomes a coordination problem. Flexprice explicitly markets to this scenario: consolidated reporting across entities, entity-specific currencies and tax handling, intercompany billing automation [5]. This is not a solved problem in most open-source billing tools.

---

Features

Based on the README, documentation, and website:

Core metering engine:

• Real-time event ingestion backed by ClickHouse [GitHub README]
• REST API + SDKs for Go, Python, and JavaScript/TypeScript [GitHub README]
• Event Debugger: real-time stream of ingested events with full payload inspection, filtering by event name, customer ID, time range [1]
• Event validation: matches event names to metered features, verifies aggregation fields, surfaces mismatches before they hit billing [2]
• Aggregation functions: Sum, Count, Max, Unique Count per event property [2]

Pricing model support:

• Usage-based pricing (pay for what you use)
• Subscription / seat-based billing
• Hybrid plans (subscription base + usage overage)
• Prepaid credit bundles with auto top-ups
• Per-customer plan overrides without engineering tickets [GitHub README][website]
• Pricing changes via no-code UI, no redeploy required [website]

Credits and entitlements:

• Prepaid credits with configurable expiration [GitHub README]
• Promotional credit grants
• Auto top-up triggers when balance drops below threshold
• Boolean and metered feature flags for access control [3]

Billing operations:

• Automatic invoice generation
• Proration handling
• Multiple currencies [3]
• Multiple environments (dev, staging, production) [GitHub README]
• Payments, CPQ, CRM, and accounting integrations via the open architecture [GitHub README]

Enterprise / multi-entity:

• Multiple environments [GitHub README]
• Multi-entity structure with consolidated reporting [5]
• Global currency support [3]

What’s not in the core self-hosted edition: The pricing page wasn’t captured in the scrape, so the exact feature split between community and commercial tiers isn’t confirmed. Assume advanced enterprise features (SSO, audit logs, fine-grained RBAC) follow the same commercial-gated pattern as comparable tools.

---

Pricing: SaaS vs self-hosted math

The Flexprice pricing page wasn’t accessible in the data for this review, so specific cloud tier prices are not available. What is known:

Flexprice self-hosted (Community Edition):

• License: AGPL-3.0 (more on the implications below)
• Infrastructure: ClickHouse + PostgreSQL + your application events pipeline
• A minimal VPS won’t cut it — ClickHouse has real memory requirements. Budget $20–40/mo on Hetzner or Contabo for a functional small-scale instance
• Cloud SaaS pricing: not captured — check https://flexprice.io/pricing directly

The comparison that matters:

Stripe Billing at scale:

• 0.5% of revenue (Stripe’s billing add-on fee)
• $500/mo at $100K MRR
• $1,000/mo at $200K MRR
• $5,000/mo at $1M MRR

Custom in-house billing:

• One developer maintaining it at 20–30% bandwidth = roughly 0.5–1 FTE-month per quarter, ongoing [website]

Flexprice self-hosted:

• $20–40/mo infrastructure
• One-time setup cost
• No per-event fees, no revenue percentage

The math is compelling past roughly $50K ARR, assuming you have engineering resources for setup. Below that threshold, Stripe’s built-in billing is simpler and the percentage fee is immaterial.

Note: AGPL-3.0 license matters here. If you’re embedding Flexprice in a product you sell, or deploying it as part of a service, AGPL requires you to release your modifications. For internal use (your own billing infrastructure) this is irrelevant. For SaaS products built on top of it, talk to a lawyer.

---

Deployment reality check

Flexprice is not a single-container deploy. The stack is:

• PostgreSQL (billing data, customers, plans)
• ClickHouse (high-volume event storage and aggregation)
• The Flexprice application itself
• Docker Compose for local/single-server deployment [GitHub README]

ClickHouse is the differentiator and the complication. It’s the reason Flexprice can handle millions of events without degrading, but it also means you’re running a column-store database that most developers haven’t operated before. ClickHouse needs a minimum 2–4GB RAM to run comfortably, and more if your event volume is serious.

What you actually need:

• A VPS with 4–8GB RAM minimum (ClickHouse alone wants 2GB)
• Docker and docker-compose
• A domain and reverse proxy for HTTPS
• An external SMTP provider for transactional email
• Familiarity with reading docker-compose logs when things misbehave

What can go sideways:

• Event name mismatches between your application and your Flexprice feature config cause silent billing errors. The Event Debugger helps — filter by customer ID, inspect exact payloads — but you need to build that verification habit [1][2]
• Aggregation field mismatches are case-sensitive. If your event sends Credits and the feature expects credits, nothing breaks loudly — it just doesn’t aggregate [2]
• ClickHouse schema migrations if you upgrade between Flexprice versions are not always straightforward for users without ClickHouse experience
• The project is approximately one year old. There will be breaking changes in the roadmap, and the migration documentation for self-hosted instances is still maturing

Realistic time estimate:

• Developer with Docker experience: 1–2 hours to a working instance
• First billing flow (events → metered feature → plan → invoice): another 2–3 hours including API integration
• Non-technical founder: this is not a deploy-yourself tool. You need a developer or a deployment service.

---

Pros and cons

Pros

• Real-time ClickHouse-backed metering. Purpose-built for high event volumes, not a Stripe meters wrapper. If you’re metering AI token consumption or API calls at scale, the architecture fits [GitHub README][3].
• No revenue-percentage pricing. Self-hosted means you pay for infrastructure, not a cut of your ARR. The math becomes compelling quickly past $100K ARR against Stripe Billing [5].
• No-code pricing changes. Product and finance teams can modify tiers, pricing, and plans without engineering tickets or redeploys [website]. This matters more than it sounds — pricing experiments blocked by deployment pipelines are a real bottleneck.
• Credit system is first-class. Prepaid credits, promotional grants, auto top-ups, expiration logic — not bolted on, but built into the core model [GitHub README][3].
• Active development. Last commit within 24 hours of this review. The project is moving fast [3].
• Multi-entity billing built in. Consolidated reporting, intercompany billing, multi-currency — not an afterthought [5].
• Event Debugger. Real-time visibility into the event ingestion pipeline with filtering and payload inspection is genuinely useful for debugging [1][2].

Cons

• AGPL-3.0, not MIT. Commercial embedding has license implications. If you’re building a product on top of Flexprice’s code, AGPL’s copyleft requirements apply [GitHub README]. For internal billing infrastructure this is fine; for resale or embedded deployment, it needs legal review.
• ClickHouse dependency raises the ops floor. This isn’t a “spin up a $5 droplet” deploy. You’re operating a columnar database most teams haven’t touched before.
• Young project with limited third-party review coverage. The primary sources for this review are Flexprice’s own documentation and their blog. There are no major independent user reviews, post-mortems, or war stories from teams running it at scale [3].
• Developer-only onboarding. A non-technical founder cannot deploy or maintain this. The no-code UI covers pricing changes after setup, but getting there requires a developer [GitHub README][2].
• Pricing page wasn’t captured. The cloud tier pricing is unknown from the available data — check the website directly before committing to the SaaS option.
• Community is early-stage. Slack community exists, 152 forks, 3,549 stars — real but small compared to Lago (9,541 stars) or established tools [3][4].

---

Who should use this / who shouldn’t

Use Flexprice if:

• You’re building an AI product that bills by tokens, API calls, or compute units and you need real-time usage enforcement — not month-end reconciliation.
• You’re past $100K ARR and Stripe’s percentage billing fee is becoming a material line item.
• Your pricing model has changed more than twice in the past year and you’re tired of engineering sprints to keep billing current.
• You have a developer who can handle Docker and a ClickHouse deployment, or budget to hire one for setup.
• You’re operating multiple legal entities and need consolidated billing reporting without spreadsheet gymnastics [5].

Skip it (consider Lago instead) if:

• You want a more battle-tested open-source billing infrastructure. Lago has 9,541 stars, has been around longer, and has more independent user reviews [3][4].
• You need a large community with documented migration guides and public post-mortems.
• Your risk tolerance for young infrastructure is low.

Skip it (stay on Stripe Billing) if:

• You’re under $50K ARR and the percentage fee is immaterial.
• Your pricing is flat subscriptions with no usage component.
• You don’t have a developer to handle deployment and ongoing maintenance.
• Your compliance or legal team restricts self-hosted infrastructure.

Skip it (use Orb or Chargebee) if:

• You need an enterprise SLA, a dedicated support team, and documented compliance certifications, and you’re willing to pay for them.
• You’re building billing for a public company or heavily regulated industry where a self-hosted OSS stack isn’t acceptable.

---

Alternatives worth considering

From the OpenAlternative listings and the project’s own positioning [3][4]:

• Lago — the most direct open-source competitor. More mature, more stars (9,541 vs 3,549), longer track record. If you want the safest open-source billing choice, Lago is the current market leader [3][4].
• OpenMeter — real-time metering with a narrower focus (events → revenue, integrates with Stripe for payments). 1,906 stars. A lighter option if you don’t need Flexprice’s full billing stack [3].
• Meteroid — newer flexible billing platform, 1,010 stars. Similar positioning but less feature-complete [4].
• Autumn — simplest option: wraps Stripe with credits and usage tracking in three functions. 2,488 stars. Much lower ops overhead, but you’re still on Stripe’s pricing [3].
• Stripe Billing — the incumbent. Zero ops overhead, largest ecosystem, well-understood. The revenue percentage fee is the trade-off [5].
• Orb — purpose-built usage-based billing, cloud-only, expensive but mature. The choice for teams that need enterprise support and don’t want to operate infrastructure [3].

For a developer team building an AI product at $50K–$500K ARR who wants to stop paying Stripe a percentage, the realistic shortlist is Flexprice vs Lago. Flexprice if real-time ClickHouse metering and credit system depth matter; Lago if community size and time-in-market matter more.

---

Bottom line

Flexprice is solving a real problem that becomes more painful the more successful your product gets: billing infrastructure designed for subscriptions doesn’t map cleanly onto usage-based or credit-based AI products, and paying a percentage of revenue to a third party for billing logic you could own gets expensive fast. The architecture — real-time ClickHouse metering, credits as a first-class primitive, no-code pricing changes — is well-designed for the problem. The trade-offs are honest: AGPL-3.0 licensing, a ClickHouse dependency that raises the ops bar, and a project young enough that you’d be one of the early production adopters. If your team has the engineering foundation to run it, and Stripe’s metering limitations or revenue percentage are active pain points, the math is worth working out. If you want a more proven open-source billing foundation with more community around it, Lago has a head start.

If the deployment overhead is the blocker, that’s exactly what unsubbed.co’s parent studio upready.dev handles for clients. One-time setup, you own the infrastructure.

---

Sources

1. Flexprice Docs — Event Debugger (official documentation). https://docs.flexprice.io/docs/event-ingestion/event-debugger
2. Flexprice Docs — Validating Events (official documentation). https://docs.flexprice.io/docs/event-ingestion/validating-events
3. OpenAlternative.co — Flexprice listing (third-party open-source directory). https://openalternative.co/flexprice
4. OpenAlternative.co — Open Source Projects tagged “Billing” (third-party directory, billing category). https://openalternative.co/tags/billing
5. Flexprice Blog — “5 Best Multi-Entity Billing Platforms For AI And SaaS In 2026” (Flexprice blog, authored by Ayush Parchure, Feb 21, 2026). https://flexprice.io/blog/best-multi-entity-billing-software-for-ai-saas

Primary sources:

• GitHub repository and README: https://github.com/flexprice/flexprice (3,549 stars, AGPL-3.0 license)
• Official website: https://flexprice.io
• Documentation: https://docs.flexprice.io