If your product charges by the API call, the token, the compute-minute, or the seat-hour, a standard subscription billing tool eventually cracks under the math. Usage-based billing software exists specifically to meter that consumption in real time and turn it into an accurate invoice — without an engineering team hand-rolling the aggregation logic every quarter.
For most SaaS and infrastructure teams building or scaling a metered pricing model in 2026, Orb is the strongest overall pick: purpose-built for usage-based billing without forcing a migration off an existing payments stack. Teams already processing payments through Stripe get there fastest with Stripe Billing's native Meters API, with Metronome — now a Stripe product — as the upgrade path once event volume gets serious.
2026 has been a consolidation year for this category: Stripe acquired Metronome in January, Salesforce acquired m3ter in July, and Zuora had already absorbed the usage-metering specialist Togai back in 2024. Two of the seven tools below aren't fully independent companies anymore, even though they still operate under their original names. For flat and tiered subscription billing beyond metered usage specifically, see our guide to billing and subscription management software.
Why You Need Usage-Based Billing Software
Bill customers for what they actually consume. Metering ties revenue directly to usage instead of a flat guess that either underprices power users or overcharges casual ones.
Process usage events at real scale. Real-time ingestion pipelines handle millions of events without a nightly batch job quietly falling behind or double-counting.
Give customers spend visibility before the invoice lands. In-product dashboards and usage alerts stop the bill from being the first time a customer learns they're over budget.
Cap runaway costs automatically. Threshold billing and usage limits catch a scraping bug or a looping AI agent before it becomes an unrecoverable write-off.
Stay audit-ready on revenue recognition. Built-in ASC 606 / IFRS 15 support keeps usage-based revenue reporting clean without a quarterly spreadsheet reconciliation project.
How We Evaluated
Every platform below was checked directly against its own pricing page, product docs, and — where one exists — its MCP server documentation, never through a review aggregator. Scoring weighs four factors: pricing transparency, real event-ingestion depth versus marketing language, AI/MCP integration maturity, and how much implementation work a mid-market team should realistically expect. Full scoring approach at our methodology page.
Editorial disclosure: none of the seven platforms in this guide paid for inclusion or placement. Rankings and editor scores reflect independent research against each vendor's own pricing, docs, and public materials.
Best 7 Usage-Based Billing Software in 2026
1. Orb
Orb doesn't treat usage-based billing as an afterthought to flat subscriptions — metered and hybrid pricing is the entire premise. The product handles three jobs: meter usage at scale, let a pricing team iterate without filing an engineering ticket, and turn the result into a bill customers can make sense of. It's a frequent reference point in API management conversations too, since so many customers are infrastructure companies billing by the call.
Pricing: Core, Advanced, and Enterprise tiers, all fully custom-quoted based on usage volume — no published starting price, and every tier routes to a sales conversation.
Top features:
Usage metering built for scale, not bolted on
Self-serve pricing simulation and iteration tools
Automated usage-to-invoice billing pipeline
Customer-facing spend and usage dashboards
Token, credit, and per-action pricing for AI products
Event ingestion via API and S3 batch upload
Pros:
Purpose-built specifically for usage-based and hybrid pricing, not a bolted-on module
Customer roster includes Vercel, Replit, Supabase, and Redis — real infrastructure-company validation
Rapid pricing iteration, demonstrated in LaunchDarkly's public case study
Cons:
Zero published pricing — every tier requires a sales conversation before you know the cost
No self-serve or trial tier for smaller teams to test the product
AI/MCP Integration: Orb markets token, credit, and per-action pricing for AI products, but no official MCP server was found on its site or docs as of August 22, 2026.
API Integration: Yes — Orb supports event ingestion via API and S3 batch upload, with SDK and developer documentation.
Cloud Based: Yes, fully cloud-hosted SaaS.
Platforms: Web dashboard, API, SDKs.
Best for: SaaS and infrastructure teams whose primary pricing model is usage-based or hybrid, and who want a platform built around that from day one.
Editor score: 4.6/5 — the cleanest pure-play usage-billing product in this list, docked slightly for offering zero pricing transparency of its own.
2. Metronome (part of Stripe)
Metronome built its reputation metering the messiest, highest-volume usage patterns in software — the kind OpenAI, Anthropic, and Nvidia generate. Stripe completed its acquisition of the company on January 14, 2026, and Metronome still runs as its own product today, with its own site, docs, and pricing, just backed by Stripe's infrastructure now.
Pricing: Startup tier — usage-based at 0.8% of billing volume plus $0.04 per 1,000 events, with native Stripe integration included. Custom/Enterprise tier — contact sales, no published rate.
Top features:
Real-time metering, pricing, billing, and reporting
Metering and data aggregation up to 100K events per second
In-product billing dashboards with spend transparency
Native Stripe integration on the Startup tier
Flexible pricing experiments and mid-cycle updates
Rate cards for complex usage-based pricing logic
Pros:
Purpose-built for the highest-volume usage patterns in the industry
Now backed directly by Stripe's infrastructure and resources
Native Stripe integration removes a redundant payments migration
Cons:
Startup-tier pricing (0.8% + $0.04 per 1,000 events) isn't the cheapest entry point
Enterprise tier pricing is fully custom and unpublished
AI/MCP Integration: No official MCP server was found on Metronome's site or docs as of August 22, 2026 — notable given how AI-heavy its customer base is.
API Integration: Yes — Metronome publishes API and webhook documentation for usage ingestion and billing operations at docs.metronome.com.
Cloud Based: Yes, fully cloud-hosted SaaS.
Platforms: Web dashboard, API, SDKs.
Best for: AI and infrastructure companies with the highest-volume usage patterns who want Stripe-backed reliability.
Editor score: 4.7/5 — the strongest technical pedigree in this list for extreme event volume, now paired with Stripe's balance sheet, held back only by startup-tier pricing that isn't the friendliest for a team just getting started.
3. Stripe Billing
Already running payments through Stripe? Stripe Billing is the path of least resistance — subscription logic, invoicing, and a customer portal live inside the platform already processing the money, with a built-in Meters API for usage tracking before anyone needs Metronome.
Pricing: 0.7% of billing volume, covering transactions processed on and off Stripe. Basic usage-based billing via the Meters API is included, with up to 100M events per month. Metronome, positioned as a separate Stripe product for advanced usage-based billing, isn't included in this price and requires its own quote.
Top features:
15+ built-in pricing models — flat, per-seat, tiered, and more
Native Meters API for usage tracking and metered pricing
100+ payment methods across 135+ currencies
Self-serve customer billing portal
AI-powered Smart Retries for payment recovery
Direct upgrade path to Metronome at scale
Pros:
Genuinely the fastest path for any team already on Stripe
Transparent, published 0.7% pricing with no hidden platform fee
Meters API covers real usage-based billing at moderate volume within the base price
Cons:
Full usage-based billing depth is pushed to the separate Metronome product
Value concentrates for businesses already committed to the Stripe ecosystem
AI/MCP Integration: Stripe Billing includes AI-driven Smart Retries for payment recovery, trained on billions of data points, but no official MCP server for billing or usage data was found on its site as of August 22, 2026.
API Integration: Yes — Stripe Billing is fully API-first, with extensive documentation, SDKs, and code samples at docs.stripe.com/billing.
Cloud Based: Yes, fully cloud-hosted SaaS.
Platforms: Web dashboard, API, SDKs, mobile SDKs.
Best for: companies already processing payments through Stripe who need usage-based billing without adding a second core vendor. For teams evaluating the payments layer itself rather than billing on top of it, our payment gateway software guide covers that comparison separately.
Editor score: 4.3/5 — excellent on-ramp with transparent pricing, but real usage-based billing depth means eventually paying for Metronome anyway.
4. Zuora
Zuora's pitch is enterprise scale, full stop — it processes usage monetization at billions of events per day for its largest customers. In 2024, it acquired Togai, a metering and rating specialist, folding that developer-focused metering engine into its own platform rather than building the capability from scratch.
Pricing: Fully custom, quote-based — no published starting price. Zuora's own materials cite billions of usage events processed daily across its customer base.
Top features:
Usage monetization at billions of events per day
Togai-derived metering and rating engine
Flexible pricing and packaging by customer segment
Any mix of recurring, usage-based, or one-time pricing
Official Zuora MCP server for AI-agent access
Milo, an AI implementation agent for quote-to-cash setup
Pros:
Genuine enterprise-scale performance, validated at billions of daily events
One of the only platforms in this list with a real, documented, official MCP server
Togai acquisition brought real developer-focused metering depth in-house
Cons:
No published pricing anywhere — every deal starts as a custom quote
Likely overkill, in both cost and implementation time, for a small or mid-market usage-billing need
AI/MCP Integration: Confirmed — Zuora publishes an official MCP server, documented at docs.zuora.com, that lets tools like Claude Desktop and Cursor query tenant data, run reports, and manage subscriptions through natural language. This is one of the more mature, genuinely documented MCP implementations in this list.
API Integration: Yes — Zuora is API-first, with native APIs, SDKs, an Extension Studio, and prebuilt connectors to Salesforce, HubSpot, and NetSuite.
Cloud Based: Yes, fully cloud-hosted SaaS.
Platforms: Web dashboard, API, SDKs.
Best for: large enterprises running complex, high-volume usage monetization who want a genuinely mature MCP implementation.
Editor score: 4.5/5 — the most credible official MCP support in this entire list, and real Togai-powered metering depth, docked for zero pricing transparency and a heavy implementation curve.
5. Chargebee
Chargebee has built a genuinely broad platform around its core billing engine — CPQ, revenue recognition, and a growth-metrics layer all live alongside subscription billing. On the usage side, it added real metered billing and AI-agent pricing support, aimed at teams monetizing tokens, credits, and agent actions.
Pricing: Flow plan pay-as-you-go runs around $400/month with a 0.80% transaction fee and no platform fee. The commit-monthly plan runs roughly $424/month with a $99 fee plus 0.65%, including 100M usage events per month. Enterprise Plus is custom-priced, demo required.
Top features:
Metered and real-time usage event ingestion
Near real-time usage aggregation at scale
Advanced metering with custom SQL configs
Usage thresholds and overage handling
Token, credit, and agent-action pricing for AI products
Official Chargebee MCP for AI-client billing access
Pros:
Real published starting pricing — rare in this category
Genuinely shipped MCP support, not a roadmap promise
Broad platform depth spanning billing, CPQ, and revenue recognition in one product
Cons:
Add-on modules like CPQ and revenue recognition require separate pricing conversations
Enterprise Plus tier pricing isn't published, and the committed plan still carries a real platform fee
AI/MCP Integration: Confirmed — Chargebee markets 'Chargebee MCP: Headless billing for any AI client' directly on its pricing page, alongside dedicated agent-pricing tooling for token, credit, and per-action billing.
API Integration: Yes — Chargebee publishes API quickstart documentation and framework-specific integrations, including Laravel Cashier and Next.js.
Cloud Based: Yes, fully cloud-hosted SaaS.
Platforms: Web dashboard, API, SDKs.
Best for: mid-market SaaS teams wanting transparent published pricing alongside real, shipped MCP-based usage billing.
Editor score: 4.5/5 — the best combination of pricing transparency and genuinely real MCP support in this list, held back only by add-on modules that scatter the total cost across separate conversations.
6. Lago
Lago takes the opposite approach from everything else here: instead of a fully managed SaaS platform, it's an open-source usage-based billing engine under an AGPLv3 license that teams can self-host. A paid Lago Cloud and Lago Premium tier exist for managed hosting or enterprise features, but the core engine is genuinely free to run yourself.
Pricing: the core open-source, self-hosted engine is free under AGPLv3. Lago Cloud and Lago Premium — managed hosting, enterprise integrations, governance — aren't publicly priced; contact sales.
Top features:
Real-time metering for flexible usage-based pricing
Hybrid plans combining flat and usage-based pricing
Entitlements and feature-gating built into billing
Dunning and revenue analytics included
Full self-hosted deployment on-prem or in a VPC
Complete public API documentation and reference
Pros:
Genuinely free, open-source core with no vendor lock-in risk
Real scale behind it — Mistral AI, Groq, PayPal, and Synthesia are named customers; Lago reports $829M in monthly invoices
Self-hosted deployment gives full control over usage data residency
Cons:
Self-hosting requires real engineering investment to run reliably in production
Cloud and Premium pricing for teams that don't want to self-host isn't published
AI/MCP Integration: Lago markets 'Lago AI' as an internal billing-intelligence feature, but this isn't an MCP server — no official MCP integration for external AI agents was found as of August 22, 2026.
API Integration: Yes — Lago provides complete API documentation, a changelog, and an active developer community.
Cloud Based: Both — self-hosted (on-prem or VPC) or Lago Cloud.
Platforms: Self-hosted via Docker/Kubernetes, Lago Cloud, API/SDKs.
Best for: cost-conscious, engineering-led teams that want full control over their billing data and are comfortable self-hosting.
Editor score: 4.2/5 — the only genuinely free option here, with real scale behind it; the honest tradeoff is that 'free' means your own team owns the operational burden.
7. m3ter (part of Salesforce)
m3ter built a metering and rating engine for consumption-based monetization, and Salesforce agreed to acquire it on June 8, 2026, completing the deal on July 1. Per Salesforce's own announcement, the plan is to fold m3ter's metering and rating capability directly into Agentforce Revenue Management — usage-based billing native to the Salesforce ecosystem.
Pricing: no pricing is published on the site — every engagement starts with a demo request, consistent with the rest of Salesforce's enterprise sales motion.
Top features:
High-volume usage data ingestion and processing
Flexible aggregation, pricing, and billing-logic configuration
Automated quote-to-cash workflows and invoicing
Self-serve billing dashboards for customers
Native integrations with Salesforce, NetSuite, and BI tools
Public REST API with full reference documentation
Pros:
Real, named enterprise customers, including Snyk, ClickHouse, and Onfido
Now backed by Salesforce's distribution and the Agentforce Revenue Management roadmap
Documented outcomes — ClickHouse reports launching a new product three times faster using m3ter
Cons:
No pricing transparency at all, not even a starting reference point
Value will likely concentrate around Salesforce-centric organizations going forward, not general-purpose usage billing
AI/MCP Integration: no mention of MCP, Model Context Protocol, or AI-agent integration was found in m3ter's public documentation as of August 22, 2026, though its new role inside Agentforce Revenue Management suggests that's the direction Salesforce is pointing it.
API Integration: Yes — m3ter publishes a public REST API with request and response samples at docs.m3ter.com.
Cloud Based: Yes, fully cloud-hosted SaaS.
Platforms: Web dashboard, API, SDKs.
Best for: enterprises already invested in the Salesforce ecosystem who want usage-based billing built into their existing revenue stack. Compared against traditional accounting software, the revenue-recognition angle here is metering-first rather than ledger-first.
Editor score: 4.0/5 — a solid metering core with real documented customer outcomes, but the least mature public pricing and content of the group, and its post-acquisition direction is still unproven.
How the 7 Usage-Based Billing Platforms Compare
| Tool | Best For | Starting Price | Standout Feature | AI-MCP Support | API Integration |
|---|---|---|---|---|---|
| Orb | Pure usage-based/hybrid pricing teams | Custom quote only | Self-serve pricing simulation | None found | Yes |
| Metronome | Highest-volume AI/infra usage | 0.8% + $0.04/1K events | 100K events/sec metering | None found | Yes |
| Stripe Billing | Teams already on Stripe | 0.7% of billing volume | Native Meters API | None found | Yes |
| Zuora | Enterprise-scale consumption billing | Custom quote only | Official Zuora MCP server | Official (confirmed) | Yes |
| Chargebee | Transparent pricing + shipped MCP | ~$400/mo + 0.80% | Chargebee MCP for AI clients | Official (confirmed) | Yes |
| Lago | Self-hosted, open-source teams | Free (self-hosted, AGPLv3) | Full self-hosted deployment | None found | Yes |
| m3ter | Salesforce-ecosystem enterprises | Custom quote only | Salesforce/Agentforce native | None found | Yes |
How to Choose Usage-Based Billing Software
Start with your event volume, not your company size. A 50-person AI startup processing a billion events a month needs Metronome- or Orb-level infrastructure; a 500-person SaaS company on light usage might not need more than Stripe's Meters API.
Check whether you're already inside a payments or CRM ecosystem. Existing Stripe customers get to usage billing fastest through Stripe Billing and Metronome; existing Salesforce shops now have a native path through m3ter.
Separate real MCP support from AI marketing language. Only Zuora and Chargebee in this list have a documented, official MCP server today — everyone else markets 'AI features' that don't actually expose billing data to an agent.
Price out the total cost, not just the headline rate. A platform's percentage fee rarely includes CPQ, revenue recognition, or enterprise support — get the full stack quoted before comparing numbers across vendors. Our pricing optimization software roundup covers the adjacent discipline of setting the prices you're then billing for.
Decide how much you're willing to operate yourself. Lago's free, self-hosted core removes vendor risk entirely, but it also moves the operational burden onto your own engineering team.
Confirm revenue recognition support before you need it. ASC 606 / IFRS 15 compliance matters the moment a metered pricing model touches audited financials — verify it's built in, not bolted on later.
Weigh implementation timeline against your launch date. Enterprise platforms like Zuora and Metronome's custom tier typically mean a longer implementation runway than a self-serve Orb or Chargebee setup.
What This Actually Costs — A Worked Example
Take a mid-size AI SaaS company processing 50 million billable usage events a month, at roughly $200,000 in monthly billing volume:
Metronome (Startup tier): 0.8% of billing volume ($1,600) plus $0.04 per 1,000 events (50,000 x $0.04 = $2,000) — roughly $3,600/month.
Stripe Billing (Meters API): 0.7% of billing volume — $1,400/month, and 50M events stays within the volume typically included before Metronome-level depth is needed.
Chargebee (commit-monthly plan): $99 platform fee plus 0.65% of billing volume ($1,300) — roughly $1,399/month, with 100M events/month included.
Orb, Zuora, and m3ter: all fully custom-quoted — expect a sales conversation before a number materializes, typically landing in a comparable or higher range once implementation and support get factored in.
Lago (self-hosted): $0 in platform fees, offset by the engineering time to run and maintain the deployment — a real cost, just not a vendor invoice.
These are illustrative estimates based on each platform's own published pricing as of August 22, 2026, not vendor quotes — actual contracts will vary with negotiated volume discounts and add-on modules.
Final Thoughts
Usage-based billing software isn't really one category anymore. It's splitting into infrastructure for teams that process events by the billion (Metronome, Orb) and usage billing bolted onto a platform you're already inside (Stripe Billing, m3ter, and increasingly Zuora after Togai). Chargebee sits in the middle: transparent pricing and a real, shipped MCP server, without demanding enterprise-scale commitment.
If AI-agent access to billing data is a real requirement today, only Zuora and Chargebee back that up with a documented MCP server — everyone else here is either silent on it or marketing adjacent AI features that don't actually expose billing data to an agent. And if another vendor line item is the real objection, Lago's free, self-hosted core is the only genuine way out of that conversation.
Two of these seven platforms changed ownership in 2026 alone, and a third did the same back in 2024 — the standalone usage-metering category may not exist in its current form much longer. For the broader subscription-billing picture beyond metered pricing, our invoicing software guide covers the rest of the stack.