# How ElevenLabs Handles Billing: A Complete Breakdown > A detailed breakdown of how ElevenLabs bills for AI voice generation using character credits, tiered subscriptions, rollover mechanics, and overage pricing - plus how to build the same model. - **Author**: Ayush Agarwal - **Published**: 2026-04-08 - **Category**: AI, Billing - **URL**: https://dodopayments.com/blogs/elevenlabs-billing-model --- ElevenLabs has become the default choice for AI-generated voice, powering everything from audiobook narration to real-time conversational agents. But what makes their business work at scale is not just the quality of their speech models. It is their billing system. Their approach combines a familiar subscription ladder with a credit-based consumption model that meters every character a user generates. Understanding this model matters if you are building an AI product with similar consumption patterns, or if you are trying to figure out how to price your own text-to-speech, image generation, or token-based platform. This post breaks down exactly how ElevenLabs structures their billing, what makes it effective, where the friction points are, and how you can build the same model using [Dodo Payments](https://dodopayments.com). ## ElevenLabs Pricing Tiers ElevenLabs offers seven tiers, starting with a free plan and scaling up through individual creator plans to business and enterprise offerings. Each tier grants a fixed monthly credit allowance alongside progressively unlocked features. | Plan | Price | Credits/Month | Approx. Minutes (Multilingual) | Key Unlocks | | ---------- | --------- | ------------- | ------------------------------ | -------------------------------------------------- | | Free | $0/mo | 10,000 | ~10 | TTS, STT, Sound Effects, Voice Design | | Starter | $5/mo | 30,000 | ~30 | Commercial license, Instant Voice Cloning | | Creator | $22/mo | 100,000 | ~100 | Professional Voice Cloning, 192kbps audio | | Pro | $99/mo | 500,000 | ~500 | 44.1kHz PCM audio output via API | | Scale | $330/mo | 2,000,000 | ~2,000 | 3 workspace seats, team collaboration | | Business | $1,320/mo | 11,000,000 | ~11,000 | Low-latency TTS (5c/min), 3 Pro Voice Clones | | Enterprise | Custom | Custom | Custom | DPA/SLAs, HIPAA BAAs, custom SSO, priority support | A few things stand out. The jump from Free to Starter is only $5, but it unlocks commercial rights and voice cloning. That is a well-placed conversion trigger. The Creator plan often runs a 50% first-month discount ($11 instead of $22), lowering the barrier to the tier where most independent creators land. The Scale and Business plans target teams and production workloads. Enterprise pricing is fully custom with negotiated SLAs, HIPAA compliance, and elevated concurrency limits. ## How ElevenLabs Billing Actually Works The billing model rests on three pillars: subscription tiers, a unified credit system, and tiered overage pricing. ### Credits as the Universal Currency Every action on ElevenLabs - text-to-speech, dubbing, voice cloning, sound effects - draws from a single pool of credits. For standard models (Multilingual v2/v3), one text character equals one credit. For Flash/Turbo models, the rate drops to between 0.5 and 1 credit per character depending on your [subscription tier](https://dodopayments.com/blogs/subscription-pricing-models). This unified pool eliminates the complexity of tracking separate quotas for different services. A creator does not need to worry about having "TTS credits" versus "dubbing credits." They have one number to watch. ### Monthly Reset with Rollover Credits reset at the start of each billing cycle. Unused credits can roll over for up to two months on active paid subscriptions. If a user downgrades or cancels, rolled-over credits expire at the end of the current billing period. The two-month rollover cap prevents infinite accumulation while still reducing the "use it or lose it" anxiety that drives [subscription fatigue](https://dodopayments.com/blogs/subscription-fatigue). A creator who has a slow month does not feel punished for staying subscribed. ### Tiered Overage Pricing Free and Starter plans block generation when credits run out. From Creator upward, users can opt into overage billing. The cost per additional unit decreases as plans get larger. | Plan | Overage Rate (per extra minute, Multilingual) | Overage Rate (Flash) | | -------- | --------------------------------------------- | -------------------- | | Free | Not available | Not available | | Starter | Not available | Not available | | Creator | ~$0.30/min | ~$0.15/min | | Pro | ~$0.24/min | ~$0.12/min | | Scale | ~$0.18/min | ~$0.09/min | | Business | ~$0.12/min | ~$0.06/min | This creates a natural upgrade incentive. If a Creator plan user regularly exceeds their 100k credits, the math quickly favors upgrading to Pro. The overage cost on Creator for 500k credits worth of generation would far exceed the $99 flat price of the Pro tier. ```mermaid flowchart TD A[New Billing Cycle] --> B[Fresh Credits + Rolled Over] B --> C[User Generates Audio] C --> D{Credits Left?} D -->|Yes| C D -->|No| E{Overage Enabled?} E -->|Yes| F[Bill at Tier Rate] F --> C E -->|No| G[Generation Blocked] D -->|Cycle Ends| H[Unused Credits Roll Over] H --> A ``` ### Per-Generation Charging Credits are deducted per generation request, not per download. If you generate the same text three times experimenting with different voice settings, you pay for three generations. ElevenLabs does offer limited free regenerations when the content and settings remain unchanged, but the default is consumption-on-generation. This is a meaningful design choice. It aligns cost with compute rather than content delivery, which reflects the actual infrastructure expense of running inference on speech models. ## What Makes This Model Effective ### Unified Credit Pool Simplifies Everything Most [AI billing platforms](https://dodopayments.com/blogs/ai-billing-platforms) struggle with the multi-product metering problem. If your platform offers TTS, dubbing, voice cloning, and sound effects, do you create separate quotas for each? ElevenLabs says no. One pool, one number. > When you are building a multi-product AI platform, the worst thing you can do is make customers manage four different usage meters. A single credit pool with different burn rates per feature gives you the flexibility of metered billing with the simplicity of a subscription. > > - Ayush Agarwal, Co-founder & CPTO at Dodo Payments This reduces cognitive load on users and simplifies the billing infrastructure. It also creates cross-selling opportunities since credits spent on trying the dubbing feature reduce the balance for TTS, nudging users toward higher tiers. ### Rollover Reduces Churn Credit rollover is an underused retention lever. According to a 2024 Zuora report, subscription businesses that offer some form of credit carryover see 15-20% lower voluntary churn than those with strict use-it-or-lose-it models. ElevenLabs caps rollover at two cycles, which is a smart middle ground. It preserves perceived value without creating a liability of unlimited stored credits on their balance sheet. For founders building [subscription models](https://dodopayments.com/blogs/one-time-vs-subscription-saas-pricing), this pattern is worth studying closely. ### Overage Pricing Drives Upgrades The decreasing overage rate at higher tiers is the most clever part of this model. It turns the upgrade decision into simple arithmetic. A user on Creator spending $22 plus $96 in overages ($0.24/min for 400 extra minutes) pays $118. The Pro plan covers the same usage for $99. The platform does not need to sell features at that point. The math sells itself. This is a pattern that works across any [metered pricing](https://dodopayments.com/blogs/metered-pricing-guide) system: make the next tier cheaper than staying where you are with overages. ## Challenges and Friction Points No billing model is perfect. Here are the areas where the ElevenLabs approach introduces complexity. ### Credit Opacity Across Models The 1:1 character-to-credit mapping only applies to standard models. Flash and Turbo models consume 0.5 to 1 credit per character depending on the plan. This variable rate makes it harder for users to predict exactly how many credits a given task will consume, especially when switching between models. For founders building similar systems, the lesson is clear: variable credit burn rates need transparent, real-time UI feedback. Users should see estimated credit cost before they click "Generate." ### No True Pay-As-You-Go ElevenLabs requires a subscription to access meaningful features. There is no option to buy a one-time credit pack without a recurring commitment. For users with sporadic, project-based needs (a podcaster who records quarterly, for example), the subscription model creates [revenue leakage](https://dodopayments.com/blogs/revenue-leakage-saas) from churn between projects. A hybrid model offering both subscriptions and one-time credit packs would capture this segment. This is one area where [flexible billing platforms](https://dodopayments.com/blogs/flexible-billing-platforms) can provide an advantage. ### Hard Caps on Lower Tiers Free and Starter users cannot opt into overages at all. When credits run out, generation stops. While this protects against bill shock, it also creates a frustrating cliff. A Starter user working on a time-sensitive project at 29,999 characters hits a wall and must either wait for the next cycle or upgrade immediately. This forced upgrade moment can feel coercive rather than organic. Offering even a small overage buffer (with appropriate [billing automation](https://dodopayments.com/blogs/billing-automation-saas) safeguards) would smooth the experience. ### Enterprise Pricing is a Black Box Enterprise plans have no published pricing, which is standard for the industry but still creates friction in procurement cycles. Buyers cannot self-serve a budget estimate, which slows deal velocity. For companies selling to enterprise customers, publishing at least a starting price or a pricing calculator reduces this friction. ## Building the Same Model with Dodo Payments The ElevenLabs billing model maps cleanly to three [Dodo Payments](https://dodopayments.com) features: [credit-based billing](https://docs.dodopayments.com/features/credit-based-billing), [usage-based metering](https://docs.dodopayments.com/features/usage-based-billing/introduction), and [subscriptions](https://docs.dodopayments.com/features/subscription). Here is how to replicate it. For a step-by-step walkthrough with full configuration details, see the [ElevenLabs billing deconstruction](https://docs.dodopayments.com/developer-resources/billing-deconstructions/elevenlabs) in the Dodo docs. ### Step 1: Define Your Credit Entitlement Create a credit entitlement called "Characters" in the Dodo dashboard. Set the credit type to Custom Unit with zero decimal precision (characters are whole numbers). Configure a 30-day expiry to match monthly billing, and enable rollover with a maximum count of 1 (credits carry forward once, then expire). This gives you the same two-cycle rollover window that ElevenLabs uses. ### Step 2: Create Tiered Subscription Products Create a subscription product for each tier and attach the Characters entitlement with tier-specific settings: | Product | Price | Credits/Cycle | Overage Enabled | Overage Rate (per 1K) | | ------- | ------- | ------------- | --------------- | --------------------- | | Free | $0/mo | 10,000 | No | - | | Starter | $5/mo | 30,000 | No | - | | Creator | $22/mo | 100,000 | Yes | $0.24 | | Pro | $99/mo | 500,000 | Yes | $0.15 | | Scale | $330/mo | 2,000,000 | Yes | $0.10 | When attaching credits to each product, uncheck **Import Default Credit Settings** to configure the per-tier overage price. This is what creates the decreasing overage rate that drives upgrades. ### Step 3: Set Up Usage Metering Create a usage meter named `tts.characters` with Sum aggregation. Link it to your Characters credit entitlement with a 1:1 meter-units-per-credit ratio. Every character your users generate maps to one credit deducted. ### Step 4: Send Usage Events from Your App Integrate event ingestion into your generation pipeline. Every time a user triggers audio generation, report the character count to Dodo: ```typescript import DodoPayments from "dodopayments"; const client = new DodoPayments({ bearerToken: process.env.DODO_PAYMENTS_API_KEY, }); async function trackGeneration( customerId: string, text: string, service: "tts" | "dubbing" | "cloning", ) { await client.usageEvents.ingest({ events: [ { event_id: `gen_${Date.now()}_${Math.random().toString(36).slice(2)}`, customer_id: customerId, event_name: "tts.characters", timestamp: new Date().toISOString(), metadata: { characters: text.length, service: service, }, }, ], }); } ``` The [event ingestion API](https://docs.dodopayments.com/features/usage-based-billing/event-ingestion) processes events asynchronously. A background worker aggregates them every minute and deducts credits from the customer's balance using FIFO ordering (oldest grants consumed first). ### Step 5: Handle Low Balance Alerts and Overages Use [webhooks](https://docs.dodopayments.com/developer-resources/webhooks) to react to credit lifecycle events: ```typescript // In your webhook handler switch (event.type) { case "credit.balance_low": // Notify user they are running low // Suggest upgrade to next tier await sendUpgradeNudge(event.data.customer_id); break; case "credit.overage_charged": // Log overage for invoice await notifyOverage(event.data.customer_id); break; case "credit.rolled_over": // Inform user of rollover await sendRolloverConfirmation(event.data.customer_id); break; } ``` Set a low balance threshold at 10% of the tier's credit quota. When a Pro user drops below 50,000 remaining credits, they get an alert - which also doubles as an upgrade prompt if they are consistently running low. ## The Upgrade Economics The most important billing lesson from ElevenLabs is how overage pricing structures the upgrade decision. Here is a concrete example: A Creator plan user ($22/month, 100k credits) who regularly generates 400k characters pays: - Base subscription: $22 - Overage: 300k extra characters at $0.24 per 1K = $72 - **Total: $94/month** The Pro plan ($99/month) includes 500k credits with no overage. The user saves nothing by staying on Creator - they are effectively paying the same amount for a worse deal. This is not accidental. It is a pricing structure that makes the "right" decision obvious. If you are building [subscription tiers with usage components](https://dodopayments.com/blogs/subscriptions-usage-based-billing-saas), designing your overage rates to cross over at the next tier's base price is one of the highest-leverage [pricing features](https://dodopayments.com/blogs/top-pricing-features-billing-tool) you can implement. ## Key Takeaways for Founders If you are building an AI product with consumption-based pricing, the ElevenLabs model offers a proven template: 1. **Use a single credit pool** across all product features. It simplifies both the user experience and your billing infrastructure. 2. **Offer limited rollover** (one or two cycles). It reduces churn without creating balance sheet liabilities. 3. **Decrease overage rates at higher tiers** to make the upgrade math obvious. 4. **Gate overages by tier** - protect casual users from surprise bills, let power users opt in. 5. **Meter at the consumption point** (generation, not download). Align cost with the actual compute expense. Whether you are building a voice AI platform, an [API-based service](https://dodopayments.com/blogs/api-monetization), or any product where users consume units of value, credit-based billing with tiered subscriptions gives you the flexibility to scale pricing with usage while keeping the user experience clean. Dodo Payments handles the full stack: [credit entitlements](https://docs.dodopayments.com/features/credit-based-billing), [usage metering](https://docs.dodopayments.com/features/usage-based-billing/introduction), [subscriptions](https://docs.dodopayments.com/features/subscription), tax compliance, and global payments in [220+ countries](https://dodopayments.com). Check the [full ElevenLabs billing deconstruction](https://docs.dodopayments.com/developer-resources/billing-deconstructions/elevenlabs) for a detailed implementation guide, or explore [pricing](https://dodopayments.com/pricing) to see how Dodo compares. ## FAQ ### How does ElevenLabs count characters for billing? For standard models (Multilingual v2/v3), one text character equals one credit. Flash and Turbo models consume between 0.5 and 1 credit per character depending on your subscription tier. Credits are deducted per generation request, not per download, so regenerating the same text counts as a new charge. ### Do unused ElevenLabs credits expire? Unused credits can roll over for up to two billing cycles on active paid subscriptions. If you downgrade or cancel, any remaining paid credits expire at the end of the current billing period. Free plan credits never roll over. ### Can you go over your ElevenLabs credit limit? On the Free and Starter plans, generation stops when credits run out. From the Creator plan upward, users can enable overage billing. Overage rates decrease at higher tiers, ranging from approximately $0.30 per extra minute on Creator down to $0.12 per extra minute on Business. ### How do you build ElevenLabs-style credit billing with Dodo Payments? Create a credit entitlement for your unit (characters, tokens, API calls), attach it to subscription products with tier-specific overage rates, set up a usage meter to track consumption, and send usage events from your application. The Dodo credit system handles rollover, expiry, FIFO deduction, and overage billing automatically. ### What is the cheapest ElevenLabs plan with commercial rights? The Starter plan at $5 per month is the entry point for commercial use. It includes 30,000 credits (approximately 30 minutes of Multilingual TTS), instant voice cloning, and a commercial license for generated audio. ## Final Thoughts ElevenLabs did not invent credit-based billing, but they executed it well for the AI voice space. The combination of a unified credit pool, limited rollover, and tiered overage pricing creates a system where users naturally migrate to higher plans as their usage grows. The model is worth studying not just for voice AI, but for any consumption-based product. If your customers burn through units of value - tokens, API calls, compute hours, storage - the same structural principles apply. The billing infrastructure to support this does not need to be custom-built. [Dodo Payments](https://dodopayments.com) gives you credit entitlements, usage metering, tiered subscriptions, and the [merchant of record](https://dodopayments.com/blogs/what-is-a-merchant-of-record) layer for tax and compliance out of the box. Build the product. Let the billing scale with it. --- - [More AI articles](https://dodopayments.com/blogs/category/ai) - [All articles](https://dodopayments.com/blogs)