ElevenLabs API Pricing Explained: Rates, Free Tier and Limits (2026)
Quick answer: The ElevenLabs API is billed in US dollars per unit of usage. Text to speech costs $0.10 per 1,000 characters with the Eleven v3 and Multilingual v2 models, or $0.05 per 1,000 characters with Flash/Turbo and v3 Conversational. A free plan includes about 10 minutes of speech per month, and paid plans run from $6 to $990 per month. Figures are from ElevenLabs' official API pricing, checked on September 27, 2026, and exclude taxes.
Is the ElevenLabs API Free?
Yes, with limits. The Free plan includes API access and 10,000 characters of v3 or Multilingual text to speech per month (roughly 10 minutes of audio), or 20,000 characters with the cheaper Flash/Turbo and v3 Conversational models. Free-plan audio cannot be used commercially; non-commercial sharing requires attribution to ElevenLabs. Paid plans include a commercial license, subject to rights, service-specific terms and exclusions for beta services. See the official publication and licensing policy before distributing generated audio.
Startups can also apply for ElevenLabs' Startup Grants Program, which offers 12 months of free usage with 33 million characters.
ElevenLabs API Rates per Product
These are the per-unit API rates ElevenLabs publishes. API usage is billed in dollars, not in the credits used by the web app.
| Product | Model | API price |
|---|---|---|
| Text to Speech | Eleven v3 | $0.10 per 1,000 characters |
| Text to Speech | Multilingual v2 | $0.10 per 1,000 characters |
| Text to Speech | v3 Conversational (low latency) | $0.05 per 1,000 characters |
| Text to Speech | Flash / Turbo | $0.05 per 1,000 characters |
| Speech to Text | Scribe v2 | $0.22 per hour |
| Speech to Text | Scribe v2 Realtime | $0.39 per hour |
| Voice agents | Speech Engine | $0.08 per minute |
| Music | Music generation | $0.15 per minute |
| Audio processing | Voice Changer, Voice Isolator | $0.12 per minute |
| Audio generation | Sound Effects | $0.12 per minute |
| Dubbing | Dubbing v1 | $0.33 per minute with watermark, $0.50 without |
| Dubbing | Dubbing v2 | $2.20 per minute |
One minute of speech is roughly 1,000 characters, so $0.10 per 1,000 characters works out to about $0.10 per minute of v3 audio and $0.05 per minute with Flash.
ElevenLabs API Pricing Plans (2026)
You can pay as you go or subscribe. A subscription gives you a fixed monthly allowance at the same per-unit rates, plus plan features such as voice cloning and more concurrent requests. The character equivalents below assume the allowance is spent entirely on that TTS model group, not shared with transcription, agents or other audio products.
| Plan | Monthly price | Annual (per month) | v3 / Multilingual characters included | Flash / Conversational characters included |
|---|---|---|---|---|
| Free | $0 | $0 | 10,000 (~10 min) | 20,000 (~20 min) |
| Starter | $6 | $5 | 60,000 (~60 min) | 120,000 (~120 min) |
| Creator | $22 ($11 first month) | $18.33 | 220,000 (~220 min) | 440,000 (~440 min) |
| Pro | $99 | $82.50 | 990,000 (~990 min) | 1,980,000 (~1,980 min) |
| Scale | $299 | $249.17 | 2,990,000 (~2,990 min) | 5,980,000 (~5,980 min) |
| Business | $990 | $825 | 9,900,000 (~9,900 min) | 19,800,000 (~19,800 min) |
| Enterprise | Custom | Custom | Custom | Custom |
Annual billing costs ten times the monthly price, which amounts to two months free.
What each plan adds
- Free: API access to text to speech, speech to text, sound effects, voice design, and music, for non-commercial use.
- Starter ($6): Commercial license and Instant Voice Cloning.
- Creator ($22): Professional Voice Cloning.
- Pro ($99): 44.1 kHz PCM audio output through the API.
- Scale ($299): 3 workspace seats and 3 Professional Voice Clones.
- Business ($990): 10 seats, 10 Professional Voice Clones, and low-latency text to speech from about 5 cents per minute.
- Enterprise: Custom terms around DPAs and SLAs, BAAs for HIPAA customers, custom SSO, elevated concurrency limits, volume discounts, and priority support.
ElevenLabs API Rate Limits and Concurrency
API concurrency is model-specific, not a single limit for every product. The official model documentation lists:
| Plan | Multilingual v2 concurrent requests | Flash concurrent requests |
|---|---|---|
| Free | 2 | 4 |
| Starter | 3 | 6 |
| Creator | 5 | 10 |
| Pro | 10 | 20 |
| Scale | 15 | 30 |
| Business | 15 | 30 |
| Enterprise | Custom | Custom |
Concurrency means requests running simultaneously, not requests per minute. These columns do not establish v3, speech-to-text, music or Agents limits. Monitor the API's concurrency response headers and use an application queue.
Per-request text limits also depend on the exact model ID: 5,000 characters for eleven_v3, 10,000 for eleven_multilingual_v2, 40,000 for eleven_flash_v2_5, and 30,000 for eleven_flash_v2.
Split longer scripts into multiple requests.
The model reference now marks Turbo models as deprecated in favor of Flash, so new integrations should select an explicit supported model ID.
Advanced Voice Models
Choosing the model is the biggest lever on API cost, because the rate halves between the two groups:
- Eleven v3 ($0.10 per 1,000 characters): The most expressive model, with emotionally rich delivery, multi-speaker dialogue, and 70+ languages.
- Multilingual v2 ($0.10 per 1,000 characters): Consistent quality for long-form audio in 29 languages.
- v3 Conversational ($0.05 per 1,000 characters): A low-latency version of v3 (about 280 ms) tuned for real-time conversation.
- Flash v2.5 ($0.05 per 1,000 characters): A low-latency model for voice agents and high-volume apps; Turbo is a deprecated predecessor.
How Much Will the ElevenLabs API Cost You?
Multiply your monthly characters by the model rate:
- A 50-minute weekly podcast in v3 is about 217,000 characters a month, or roughly $22. Creator's 220,000 included characters cover this estimate if the allowance is used only for TTS.
- A voice agent speaking 5,000 minutes a month on Flash uses about 5 million characters, or roughly $250 for TTS alone. Scale covers up to 5,980,000 Flash characters; agent-platform, transcription and LLM charges are separate.
- An audiobook of 500,000 characters in Multilingual v2 costs about $50. You can pay as you go if it is a one-off, or use a month of Pro.
Unused subscription credits roll over for up to two months while your paid plan stays active. Downgrading or cancelling forfeits them at the end of the billing cycle. Check the billing terms for your account; web-app credits and API dollar allowances are not interchangeable units.
Pros and Cons of ElevenLabs API Pricing for Developers
Pros
- Simple, published per-unit rates. Every product has a dollar price per character, minute, or hour, so estimating cost is simple arithmetic.
- Pay-as-you-go option. You can use the API without a subscription and pay only for what you use.
- A cheap tier for real-time work. Flash/Turbo and v3 Conversational cost half as much as the premium models while staying low latency.
- Voice cloning included. Instant cloning starts at $6 per month and professional cloning at $22.
- Rollover. Paid plans carry unused credits for up to two months.
Cons
- Compare bulk costs. Price the same script and required output quality with alternative TTS services before committing to high-volume narration.
- No commercial use on Free. Non-commercial publication requires attribution; commercial use needs an eligible paid license.
- Model-specific concurrency caps. Free supports 2 simultaneous Multilingual v2 requests or 4 Flash requests, so multi-user apps need queuing.
- Features tied to tiers. Professional voice clones, seats, and PCM output require moving up plans even if your volume is low.
- SLAs only on Enterprise. Formal uptime and data-processing commitments require a custom contract.
Conclusion: Is ElevenLabs API Pricing Worth It?
For products where voice quality is the feature, such as narration, character voices, dubbing, or branded voice agents, compare ElevenLabs' $0.05-$0.10 per 1,000 characters with your output-quality requirements. Start on Free to test voices, then check licensing and concurrency before production use. For plain bulk narration, compare total costs with other services in our text-to-speech guides. For implementation ideas, see n8n and ElevenLabs voice automation.
Frequently Asked Questions
How much does one million characters cost?
At the listed base rates, one million TTS characters cost $100 with v3 or Multilingual v2, or $50 with Flash or v3 Conversational, before taxes or negotiated discounts. This is a usage estimate, not an additional subscription fee.
Are API credits the same as characters?
No. Characters measure the TTS input, while credits or dollar allowances measure an account's budget. The number of characters that budget buys depends on the model and any other products consuming it.
Can I publish free-plan audio commercially?
No. Free-plan audio requires attribution when shared non-commercially and does not gain a commercial license just because you subscribe later. Read the linked licensing policy, including the beta-service exclusions.
Is Business limited to 25 API requests at once?
Not in the documented TTS API limits. Business supports 15 concurrent Multilingual v2 requests or 30 Flash requests. A creative-product plan table is not a substitute for the model-specific API documentation.
Sources
- ElevenLabs API pricing: per-product rates, included characters per plan, and model limits
- ElevenLabs plans and pricing: plan prices, features, concurrency, rollover, and annual billing
- ElevenLabs model reference: API concurrency, exact model IDs, request-size limits and deprecations
- Publication and licensing policy: commercial rights, attribution and beta exclusions