How Much Does ElevenLabs Agents Cost in 2026? Plans, Per-Minute Rates and Real Bills
ElevenLabs Agents charges $0.08 for every minute of call time on every plan, from the $6 Starter to the $990 Business, and $0.003 per text message. The plan sets your prepaid minutes and how many calls can run at once, while the language model and the phone line are billed on top.
Key takeaways
- ElevenLabs Agents charges $0.08 a minute for call time on every plan, from the $6 Starter to the $990 Business, whether the minute comes from the plan's included allowance or is bought extra.
- Text messages cost $0.003 each, burst minutes above a plan's concurrency limit cost $0.16 a minute, and yearly billing cuts the effective included-minute rate to about $0.067.
- The LLM is billed per minute on top of the platform rate, ranging from $0.0005 a minute on GPT-5 Nano to $0.0446 on GPT-5.5.
- Voice calls get a 95% discount for silences longer than 10 seconds, but billing runs on total connection duration, from when the call starts to when it ends or the window closes.
- A 1,000-minute support line on a toll-free number costs $105.35 to $130.05 a month depending on the LLM, rising to $2,066.15 to $2,560.15 a month at 20,000 minutes.
ElevenLabs Agents costs $0.08 per call minute on every plan, from the free tier with 15 minutes to the $990 Business plan with 12,375, and the language model and your phone carrier are billed on top. Plans differ in how many calls can run at once, not in the price of a minute.
How much does each ElevenLabs Agents plan cost?
| Plan | Monthly | Billed yearly | Call minutes included | Concurrent calls | Also includes |
|---|---|---|---|---|---|
| Free | $0 | $0 | 15 | 4 | Workflow builder, knowledge base, multilingual, widget |
| Starter | $6 | $5 a month | 75 | 6 | Text messages, commercial license |
| Creator | $22 ($11 first month) | $18.33 a month | 275 | 10 | Additional minutes |
| Pro | $99 | $82.50 a month | 1,238 | 20 | |
| Scale | $299 | $249.17 a month | 3,738 | 30 | 3 workspace seats |
| Business | $990 | $825 a month | 12,375 | 40 | 10 workspace seats, 10 professional voice clones |
| Enterprise | Custom | Custom | Custom | Higher | HIPAA BAAs, custom SSO, DPA and SLA terms, "significant discounts at scale" |
On every plan: extra call minutes $0.08, text messages $0.003, burst minutes $0.16, the LLM "based on usage and varies by model", and telephony "at cost". Yearly billing is two months free.
Prices checked September 19, 2026 on ElevenLabs' ElevenAgents pricing page, in US dollars. ElevenLabs shows rupees to visitors from India, so a fetch from there won't match this table.
What the product does is in our ElevenLabs Agents guide.
Why does every plan work out to $0.08 a minute?
Divide each plan's price by its included minutes. Starter is $6 for 75 minutes, Creator $22 for 275, Pro $99 for 1,238, Scale $299 for 3,738, Business $990 for 12,375. Every one comes to $0.080 a minute, the same as an extra minute bought on top.
That changes how to read the table. A bigger plan doesn't buy cheaper minutes on monthly billing. It buys concurrency (6 calls at once on Starter, 40 on Business), workspace seats and voice clones, and it commits you to a minimum spend. The only volume discount on the page is yearly billing, which cuts the effective rate on included minutes to about $0.067.
So the rule for choosing a plan is concurrency first. Estimate your peak simultaneous calls, pick the smallest plan that covers them, and buy the remaining minutes at $0.08. The ElevenAgents FAQ says plans are "billed by call minutes, not by the shared credit pool" that ElevenLabs' voice-generation plans use.
Two measurement details work in your favor and one against. ElevenLabs' help center says voice calls get a 95% discount for silences longer than 10 seconds, so a caller put on hold doesn't cost full rate. It also says billing runs on "the connection duration", from when the call starts to when it ends or the window closes, so the billed time can be longer than the conversation.
What happens when you use up your included minutes?
For new self-serve customers, extra minutes come from a prepaid balance, not an overage bill. ElevenLabs' Pay As You Go docs describe Top Ups of at least $5, available on every self-serve tier including Free. Your plan's minutes are used first, then the balance.
The consequence for a phone line: "If your balance hits $0 and Auto Top Up is off (or fails), your service will pause immediately. This is by design. PAYG is a prepaid model with no overage charges." A support line that goes silent mid-afternoon because a card declined is worse than an overage invoice. Turn on Auto Top Up and keep a buffer. Unused Top Up balance expires after 12 months.
Burst pricing is the other lever. When enabled, agents can take up to three times your plan's concurrent calls, and the calls above your limit cost $0.16 a minute, double the standard rate. It's cheaper than upgrading for an occasional spike and more expensive for a daily one.
How much do the LLM and the phone line add?
ElevenLabs passes both through. Its pricing calculator shows an estimated LLM cost per minute for each model, which we read by selecting each one on September 19, 2026:
| Model | LLM cost per call minute |
|---|---|
| GPT-5 Nano | $0.0005 |
| Gemini 2.5 Flash | $0.0012 |
| GPT-5 Mini | $0.0024 |
| Claude Haiku 4.5 | $0.0086 |
| GPT-5 | $0.0119 |
| Gemini 3.1 Pro Preview | $0.0178 |
| Claude Sonnet 4.6 | $0.0259 |
| Claude Opus 4.7 | $0.0431 |
| GPT-5.5 | $0.0446 |
On a fast, cheap model the LLM adds about 1.5% to the $0.08 platform minute. On GPT-5.5 it adds 56%. Inside the product, the same estimates appear in the agent's model picker, next to each model's typical latency, and a "Detailed costs" button estimates an agent's LLM spend from its past calls.
The docs screenshot and the calculator differ slightly (Claude Haiku 4.5 is ~$0.0082 in the docs image, $0.0086 on the calculator). Use the calculator for current figures.
Telephony comes from your carrier or SIP trunk at its own rates. On Twilio's US price list, receiving a call costs $0.0085 a minute on a local number and $0.0220 on a toll-free number, plus $1.15 or $2.15 a month for the number. A US support line on a toll-free number adds more than a quarter to the $0.08 platform minute.
What does ElevenLabs Agents cost at 1,000, 5,000 and 20,000 minutes?
Voice: an inbound support line on a Twilio toll-free number, monthly billing, the cheapest plan that covers the minutes (plan price plus $0.08 minutes, since every plan prices a minute the same), with two model choices. Chat: text-only conversations of 8 messages each, which need at least the Starter plan for text messages.
| Monthly volume | Plan and minutes | Platform | LLM: Gemini 2.5 Flash / Claude Sonnet 4.6 | Toll-free telephony | Total |
|---|---|---|---|---|---|
| 1,000 voice minutes (a small line) | Creator $22 + 725 extra minutes × $0.08 ($58) | $80 | $1.20 / $25.90 | 1,000 × $0.022 + $2.15 = $24.15 | $105.35 / $130.05 |
| 5,000 voice minutes (a mid-sized line) | Pro $99 + 3,762 × $0.08 ($300.96), or Scale $299 + 1,262 × $0.08 ($100.96) | $400 | $6 / $129.50 | $110 + $2.15 = $112.15 | $518.15 / $641.65 |
| 20,000 voice minutes (a large line) | Business $990 + 7,625 × $0.08 ($610) | $1,600 | $24 / $518 | $440 + $2.15 = $442.15 | $2,066.15 / $2,560.15 |
| 1,000 chats | Starter $6 + 8,000 messages × $0.003 ($24) | $30 | Not published per message | n/a | $30 before the LLM |
| 5,000 chats | Starter $6 + 40,000 × $0.003 ($120) | $126 | Not published per message | n/a | $126 before the LLM |
| 20,000 chats | Starter $6 + 160,000 × $0.003 ($480) | $486 | Not published per message | n/a | $486 before the LLM |
At 5,000 minutes, Pro and Scale cost the same $400. Scale's 30 concurrent calls against Pro's 20 are free, so take Scale. At 1,000 minutes Starter plus extra minutes also comes to $80, but Creator gives 10 concurrent calls to Starter's 6 for the same money. At 20,000 minutes, yearly billing drops Business to $825 a month and the platform line to $1,435.
The chat rows are floors. ElevenLabs charges the LLM per conversation for text too and shows the estimate in the product, but doesn't publish a per-message figure. The silence discount would lower the voice rows for calls with long holds.
What drives an ElevenLabs Agents bill up?
The model. From $0.0005 to $0.0446 a minute, so the choice can add almost nothing or more than half again to every minute.
Telephony. Toll-free inbound on Twilio adds $0.022 a minute; a local number adds $0.0085.
Concurrency. Peak simultaneous calls decide your plan. Burst calls above the limit cost $0.16 a minute.
Connection time. Billing runs from connect to hang-up, including setup time and any window left open.
Seats and voices. Workspace seats appear on the plan cards from Scale up, and professional voice clones from Business.
Downgrades. Paid credits expire when you downgrade or cancel, at the end of the cycle.
Who does per-minute billing favor?
A flat $0.08 minute means ElevenLabs earns the same on a two-minute call that resolves a question and a nine-minute call that loops before handing off to a human. The incentive runs toward more minutes, and nothing in the price rewards a shorter call. The plans themselves are minute prepayments, so the real upsell is concurrency: a team that grows past 20 simultaneous calls has to move from Pro to Scale whether or not it needs more minutes.
Passing the LLM and telephony through at cost is the part that favors the buyer. You can pick a model for price and latency and see the rate before you commit, and you can bring your own carrier. The flip side is that the headline number is never the whole bill.
What do users say about ElevenLabs pricing and billing?
ElevenLabs' public reviews cover the whole company, mostly its voice-generation tools, not only Agents. Trustpilot rates elevenlabs.io 3.0 out of 5 from 1,150 reviews, and G2's whole-company listing rates it 4.5 out of 5 from 1,213 reviews (both read September 19, 2026); neither breaks out Agents specifically.
- A reviewer in Colombia wrote on Trustpilot (trustpilot.com/review/elevenlabs.io, newest first) on September 18, 2026: "It feels like the service is designed to drain your credits before giving you acceptable quality."
- A reviewer in India, September 13, 2026: "Whenever you face a critical bug, billing issue, or credit failure, their automated system or support agents simply give you the runaround".
- A reviewer in the UK, September 2, 2026, described being "charged an annual subscription on day 4" after thinking they were on the free tier.
The recurring complaints are about credits, billing surprises and slow support, which is worth weighing on a product billed by the minute. Positive reviews on the same page praise individual support agents who fixed a problem quickly.
Vendr, which negotiates software contracts for buyers, puts the median ElevenLabs contract at $64,735 a year, from $24,668 to $132,000, "based on data from 40 purchases". That covers all ElevenLabs products and enterprise deals, not only Agents. ElevenLabs doesn't sell through AWS Marketplace directly; a search there turns up only a partner's Amazon Connect integration.
How does ElevenLabs Agents compare with Retell, Vapi and Google at 5,000 minutes?
Same volume for each: 5,000 minutes a month, agent platform and LLM only, before telephony.
| Platform | Billing unit | Published price | 5,000 minutes | Notes |
|---|---|---|---|---|
| ElevenLabs Agents | Call minute | $0.08 a minute on every plan, LLM at cost | $406 (Gemini 2.5 Flash) to $529.50 (Claude Sonnet 4.6) | Scale plan, 30 concurrent calls |
| Retell | Minute, by component | Voice infrastructure $0.055, platform voices $0.015 (ElevenLabs voices $0.040), GPT 5 mini $0.008 | $390, or $515 with ElevenLabs voices | Telephony priced separately |
| Vapi | Minute | $0.05 hosting, models and voices at provider cost | About $410 to $645 | Vapi's own estimate is $82 to $129 for 1,000 minutes; we scaled it by 5 |
| Google Conversational Agents | Second of audio per turn | $0.001 a second (Flows), $0.002 (Playbooks) | $300 (Flows) to $600 (Playbooks) | Bills listening and speaking time per turn |
Sources: Retell pricing, Vapi pricing, Google Conversational Agents pricing, checked September 19, 2026.
At this volume ElevenLabs, Retell and Vapi land within about $20 of each other on a cheap model, and Google's Flows is about $100 cheaper, so price isn't what separates the first three. ElevenLabs' case is its own voices and the simplest pricing page of the four; Retell and Vapi let you mix components, and Retell charges $0.040 a minute to use ElevenLabs voices on its platform. Our Retell, Vapi, Google Customer Engagement Suite and Parloa guides cover the rest of each product.
If your support volume lives in email and chat tickets in Zendesk, Freshdesk, Gorgias, Front, HubSpot or Intercom, a per-minute voice platform prices the wrong unit. Macha fits teams already running one of those help desks: it's an AI agent layer on your help desk, from $299 a month for 750 tickets (about $0.40 a ticket), one charge per ticket however many replies it takes, with setup and monitoring by the Macha team. It connects to ElevenLabs when an agent needs a voice; our Macha and ElevenLabs integration post shows how. Our overview of AI agents for customer service covers where voice platforms and help desk agents each fit.
How we researched this
We checked ElevenLabs' ElevenAgents pricing page and calculator (reading the LLM rate for each model), its help center and Pay As You Go docs, Twilio's US voice prices, Retell's, Vapi's and Google's pricing pages, Vendr and AWS Marketplace on September 19, 2026, and read Trustpilot in a browser the same day. We did not create or use an ElevenLabs account for this post, so every cost is built from published rates, and the chat profile (8 messages a conversation) and the toll-free line are our assumptions. We also read G2's whole-company listing (4.5 from 1,213 reviews), and we found no ElevenLabs page on Gartner Peer Insights.
Frequently asked questions
How much does ElevenLabs Agents cost per minute? $0.08 a minute on every plan, whether the minute is included in the plan or bought extra, plus the LLM (from $0.0005 to about $0.045 a minute depending on the model) and your telephony. Burst minutes above your concurrency limit cost $0.16.
How much does ElevenLabs Agents cost per month? Plans run from $0 (15 minutes) and $6 (75 minutes) to $990 (12,375 minutes), or $5 to $825 a month billed yearly. In our model a support line costs about $105 to $130 a month at 1,000 minutes and $2,066 to $2,560 at 20,000, including toll-free telephony.
Is there an ElevenLabs Agents pricing calculator? Yes. The ElevenAgents pricing page has one: pick subscription or pay as you go, set monthly minutes and choose a model, and it recommends a plan and shows the LLM cost. It leaves out telephony, which you add from your carrier's rates.
What happens if I use more minutes than my plan includes? New self-serve customers draw extra minutes from a prepaid Pay As You Go balance at $0.08 a minute. If the balance reaches $0 with Auto Top Up off or failing, service pauses until you top up. Older subscriptions with usage-based billing and Enterprise contracts are billed for overage instead.
Does ElevenLabs charge for text chat with an agent? Yes, $0.003 per text message plus the LLM, and text messages need at least the Starter plan. A chat of 8 messages costs $0.024 before the model.
Is ElevenLabs worth the money? For voice quality and a published per-minute price, it compares well: at 5,000 minutes it lands within about $20 of Retell and Vapi on a cheap model. Trustpilot reviewers rate the company 3.0 out of 5 across all products, with recurring complaints about credits, billing and support speed.
How much does it cost to use an AI voice agent? On per-minute platforms in September 2026, about $0.06 to $0.13 a minute for the platform and model, plus telephony: $0.08 on ElevenLabs before the model, from $0.078 on Retell with GPT 5 mini, and $0.06 to $0.12 on Google's Conversational Agents. Enterprise contact-center products are usually quoted per contract instead.
Does ElevenLabs have enterprise pricing? Yes, custom pricing with higher concurrency, BAAs for HIPAA, custom SSO and "significant discounts at scale". Vendr's median ElevenLabs contract across all products is $64,735 a year. Speech-to-text (Scribe) and text-to-speech API usage are priced separately on ElevenLabs' API pricing page.
If your team's volume is tickets in a help desk rather than calls, see Macha's pricing page or start with $50 of free usage on your own tickets.
Sources: ElevenAgents pricing · How much does ElevenAgents cost? (help center) · Pay As You Go (docs) · Twilio US voice pricing · Retell pricing · Vapi pricing · Google Conversational Agents pricing · Vendr: ElevenLabs · Trustpilot: elevenlabs.io (trustpilot.com/review/elevenlabs.io; loads in a browser, blocks scripted link checks) · Macha ElevenLabs integration
Simple per-ticket AI pricing
One thread with one person is one charge, however many replies it takes.
Intercom
Shopify
Stripe
Slack
Notion
Google Workspace
Confluence

