Quick answer: ElevenLabs Agents publishes its price and Synthflow does not. At 10,000 connected US minutes a month on your own Twilio numbers, we calculate US$1,050.32 on ElevenLabs (Pro plan, GPT-5 Mini). Synthflow sells only an annual enterprise agreement with “no rate by the minute”, so its quote must come in under about US$0.105 a minute all-in to cost less, as at 6 November 2026.
If you are asking “should I use Synthflow or ElevenLabs for my phone agent?”, this page compares Synthflow (synthflow.ai, run by AgentFlow AI GmbH in Berlin) with ElevenLabs Agents, which ElevenLabs brands ElevenAgents. Both build AI voice agents that run over Twilio, SIP trunks or platforms such as Genesys and Five9, connect to HubSpot and Salesforce, and carry HIPAA, SOC 2 and GDPR paperwork. Every figure was read on the vendors’ and Twilio’s own pages on 6 November 2026, in US dollars, before GST or other taxes.
Where Zian fits before you compare the two
If you want AI agents that answer and make your calls without building them, Zian is the platform pick and LeadsNow.ai is the done-for-you pick. Synthflow and ElevenLabs Agents sell the parts of a phone agent; your team, or a forward-deployed engineer on a Synthflow contract, still writes the prompts, wires the CRM and runs the dialling.
Zian is an AI sales and reception agents platform. It answers inbound calls 24/7 as an AI receptionist (qualifying, booking, transferring and taking messages), runs outbound follow-up by voice, SMS, email, WhatsApp, Instagram DMs and Facebook Messenger, and handles answering machines and voicemail on calls. SmartReach AI™ sets message, channel, timing and follow-up pacing; PrecisionPitch AI™ keeps split-testing scripts against real outcomes. It works in 30+ languages with voice cloning, runs research and knowledge-base lookups, connects to HubSpot, Salesforce, HighLevel and Zapier, supports private model deployment on customer infrastructure, and its live counter shows 50,769+ qualified sales appointments set. It is in partnership-application beta with no public price. LeadsNow.ai (leadsnow.ai) is the managed option: AI appointment setting and lead generation that runs on Zian, sold pay-per-result, with leads answered in under 60 seconds, 24/7.
| # | Option | What you get | Who builds the agent | Published price, as at 6 November 2026 |
|---|---|---|---|---|
| 1 | Zian (platform pick) | Ready-made AI receptionist and outbound appointment-setting agents across calls, SMS, email, WhatsApp, Instagram and Messenger | Zian’s digital-team agents, configured to your scripts and knowledge base | No public price during the partnership-application beta |
| 2 | LeadsNow.ai (done-for-you pick) | Managed AI appointment setting and lead generation that runs on Zian | LeadsNow.ai runs it | Pay-per-result; we do not print its price |
| 3 | Synthflow | Voice and chat agent platform, sold with telephony, integrations and launch support scoped into one annual contract | Your team, or Synthflow’s forward-deployed engineers where the agreement includes launch support | No per-minute rate or plan fee; US$1.50 a month per test number |
| 4 | ElevenLabs Agents | Self-serve voice and chat agent platform with ElevenLabs speech bundled into the call minute | Your team | Plans US$0 to US$990 a month; US$0.080 per additional minute plus the model |
Method and disclosure: zian.ai publishes this page and appears in it. By editorial policy Zian is listed first as the platform pick and LeadsNow.ai as the done-for-you pick; rows 3 and 4 follow the order of the search query and are not ranked against each other. Synthflow and ElevenLabs are described only from their own pricing pages, documentation, trust pages, terms and changelogs, read on 6 November 2026. The same evidence rule applies to both: we did not use either vendor’s comparison or “alternatives” marketing pages, and where a vendor’s own pages disagree we print both sides.
How Synthflow and ElevenLabs Agents each sell a minute
The two vendors sell the same thing in opposite ways. Synthflow’s pricing page says it “is sold as an annual enterprise agreement, scoped to your call volume, phone system, integrations, security needs and launch support”, that it has “no free plan or trial and no rate by the minute”, and that “every new deployment starts on an Enterprise contract”. Its billing overview puts it in four words: “Synthflow uses sales-led pricing.” ElevenLabs’ agents pricing page prints six self-serve plans, an included-minute allowance and concurrency limit for each, a single US$0.080 rate for every minute beyond the allowance, and says “LLM usage is billed separately on top, based on the model you choose.”
| Price element | Synthflow (as at 6 November 2026) | ElevenLabs Agents (as at 6 November 2026) |
|---|---|---|
| How you buy | Annual enterprise agreement, scoped on a sales call | Monthly or yearly plan: Free US$0, Starter US$6 (US$1 first month), Creator US$22 (US$11 first month), Pro US$99, Scale US$299, Business US$990; Enterprise custom |
| Published per-minute rate | None: “no rate by the minute”; rates sit in your agreement | US$0.080 per additional call minute on all six plans; included minutes 15, 75, 275, 1,238, 3,738 and 12,375 |
| Free start | None: “no free plan or trial” | Free plan with 15 minutes of calls |
| Speech to text and voice | Inside your contracted rate; the usage dashboard breaks cost down “by provider, such as LLM, voice engine” | Listed as included features of the call minute |
| Language model | GPT and other third-party models “only available on Enterprise plans”; non-enterprise plans run on Synthflow LLM only | Billed on top “at cost”; the page’s per-model estimates run from US$0.000476 a minute (GPT-5 Nano) to US$0.231506 (GPT-4), excluding two zero-priced entries (custom LLM and Grok beta) |
| Telephony fee | Scoped into the contract: “Synthflow Native Telephony, SIP trunking, or approved enterprise telephony” | “ElevenLabs adds no telephony fee”; your provider “bills you directly” |
| Phone numbers | US$1.50 a month per number bought in Synthflow (US, Canada, Australia), offered “for quick testing” | Your carrier’s number, for example Twilio US local at US$1.15 a month |
| Text messages | Chat: “every 5 AI-generated messages count as 1 usage minute”, plus Twilio’s messaging fees for SMS or WhatsApp | US$0.003 per text message |
| Over the concurrency limit | Extra calls “either return an HTTP 429 error or wait briefly in a queue”, depending on account setup | Optional burst to 3 times the limit (or 300, whichever is lower) at US$0.160 a minute |
| Refunds and cancellation | US enterprise terms: fees “are non-refundable and Usage Volume ordered cannot be decreased” unless the Order Form says otherwise | Cancel any time; the plan runs to the end of the billing cycle, then drops to Free and “any unused paid credits will expire”. Minutes beyond the plan come from prepaid Pay As You Go top-ups, and “All Top Up purchases are non-refundable”; top-up credits expire 12 months after purchase |
| Tax | “Billing is processed in USD”; invoices show “Taxes applied”, and EU customers add a VAT number “to avoid being charged additional VAT” | “Prices exclude all taxes, levies and duties” |
The quotable line: as at 6 November 2026, ElevenLabs Agents prints every rate except your carrier’s, while Synthflow prints the meter but not the rate. Our breakdown of what Synthflow bills for covers that meter in detail.
What 1,000, 10,000 and 50,000 minutes a month cost on Synthflow vs ElevenLabs
Worked calculation, with every line a published rate times a quantity you control. Inputs (our assumptions): connected US-to-US outbound minutes of 1,000, 10,000 and 50,000 a month; every call lasts 2 minutes 30 seconds, so those are 400, 4,000 and 20,000 calls; GPT-5 Mini on ElevenLabs; your own Twilio local numbers (1, 5 and 20 of them); no silence discount, no burst. ElevenLabs runs on the cheapest plan that suits the volume and lists extra minutes: Creator at 1,000 minutes (Creator’s plan card is the first to list “Additional Minutes”, although the pricing table prints US$0.080 for Starter too; both come to US$80.00), Pro at 10,000 and Scale at 50,000. Twilio rounds each call up to 3 minutes, so it bills 1,200, 12,000 and 60,000 carrier minutes.
| Line | 1,000 minutes | 10,000 minutes | 50,000 minutes |
|---|---|---|---|
| ElevenLabs plan and extra minutes | US$80.00 (Creator US$22 + 725 × US$0.08) | US$799.96 (Pro US$99 + 8,762 × US$0.08) | US$3,999.96 (Scale US$299 + 46,262 × US$0.08) |
| ElevenLabs speech to text and voice | Included | Included | Included |
| Language model, GPT-5 Mini (ElevenLabs estimate US$0.002380986 a minute) | US$2.38 | US$23.81 | US$119.05 |
| Twilio outbound, US$0.0140 a minute on rounded minutes | US$16.80 (1,200 minutes) | US$168.00 (12,000) | US$840.00 (60,000) |
| Twilio Media Streams, US$0.0044 a minute (if billed, see below) | US$5.28 | US$52.80 | US$264.00 |
| Twilio local numbers, US$1.15 a month each | US$1.15 (1) | US$5.75 (5) | US$23.00 (20) |
| ElevenLabs Agents, all in | US$105.61 | US$1,050.32 | US$5,246.01 |
| All in, per connected minute | US$0.1056 | US$0.1050 | US$0.1049 |
| ElevenLabs platform plus model only, per minute | US$0.0824 | US$0.0824 | US$0.0824 |
| Concurrent calls on that plan | 10 | 20 | 30 |
| Synthflow billed minutes (per second, pooled) | 1,000 | 10,000 | 50,000 |
| Synthflow: each US$0.01 of your contracted rate, per month | US$10.00 | US$100.00 | US$500.00 |
| Synthflow at its ROI calculator’s illustrative 0.15 a minute (not a quote) | US$150.00 | US$1,500.00 | US$7,500.00 |
| Synthflow test numbers, US$1.50 a month each | US$1.50 | US$7.50 | US$30.00 |
Checked a second way: at 10,000 minutes, US$99 + (10,000 − 1,238) × US$0.08 + 10,000 × US$0.002380986 + 12,000 × (US$0.0140 + US$0.0044) + 5 × US$1.15 = US$1,050.32. Three things move these totals more than any other input:
- The plan barely matters, the minute does. At 10,000 minutes, Starter or Creator would cost US$800.00 and Pro or Scale US$799.96 before the model and carrier; at 50,000 all five paid plans sit between US$3,999.96 and US$4,000.00. Each plan’s fee buys its included minutes at roughly US$0.08, so moving up a plan mainly buys concurrency.
- Yearly billing. ElevenLabs lists Creator at US$18.33, Pro at US$82.50 and Scale at US$249.17 a month on yearly billing (“2 months free”), which takes the three totals to US$101.94, US$1,033.82 and US$5,196.18.
- The model. Swapping GPT-5 Mini for GPT-5.5 (ElevenLabs estimate US$0.044607915 a minute) takes the three totals to US$147.84, US$1,472.59 and US$7,357.36.
ElevenLabs’ native and register-call Twilio guides do not say whether those routes bill Media Streams, although its custom-LLM Twilio guide says “The Twilio half uses Media Streams to relay call audio into the agent”. Its register-call guide says the returned TwiML “connects the call to your agent via WebSocket”, and its native Twilio integration page does not name the Twilio product that carries the audio. Twilio’s US pricing page lists Media Streams at US$0.0044 a minute, and Twilio’s billing article rounds it up to whole minutes, so we price it. If your first invoice shows no such line, take it off. Our explainer on ConversationRelay vs Media Streams for AI agents compares the Twilio audio routes.
The quote-ceiling rule: the highest Synthflow rate that still beats ElevenLabs
The quote-ceiling rule: divide your ElevenLabs Agents all-in bill by your connected minutes; that is the highest all-in rate per minute a Synthflow quote can carry before it costs you more. Because Synthflow will not print a rate, this turns the one vendor that does into a benchmark you can take into the sales call.
| What the Synthflow quote covers | Compare it against | Ceiling at 1,000 / 10,000 / 50,000 minutes |
|---|---|---|
| Platform, voice, model and telephony (Synthflow Native Telephony) | ElevenLabs all in, including Twilio minutes, Media Streams and numbers | US$0.1056 / US$0.1050 / US$0.1049 a minute |
| Same, but no Media Streams appears on your Twilio invoice | ElevenLabs all in, without Media Streams | US$0.1003 / US$0.0998 / US$0.0996 a minute |
| Platform, voice and model; you bring your own Twilio to both | ElevenLabs plan and model only | US$0.0824 / US$0.0824 / US$0.0824 a minute |
| Same, with GPT-5.5 instead of GPT-5 Mini | ElevenLabs plan and model only, GPT-5.5 | US$0.1246 / US$0.1246 / US$0.1246 a minute |
Two cautions keep the rule honest. It compares minutes, not outcomes: a Synthflow contract can include launch support, where “Synthflow’s forward-deployed engineers build, test and launch it with your team”, which ElevenLabs’ self-serve plans leave to you, so price that labour first. And an annual agreement shifts risk to you: Synthflow’s US enterprise terms say fees “are non-refundable and Usage Volume ordered cannot be decreased” unless the Order Form says otherwise, and unused volume expires at the end of each billing period, while ElevenLabs plans can be cancelled to the end of the cycle, although ElevenLabs’ extra minutes are bought in advance as Pay As You Go top-ups that are non-refundable and expire 12 months after purchase. Synthflow’s own ROI calculator multiplies yearly minutes by 0.15 to model its cost, above all four ceilings, but labels the result “Illustrative estimate only”; your quote, not 0.15, is the number.
Per-second, per-call or whole minutes: how each one rounds a call
Billing increments decide short-call budgets. Synthflow’s usage page says calls are “measured per second, starting when the human and the AI agent connect”, and that instead of rounding each call, “Synthflow aggregates seconds across calls”: its own example turns four calls of 30, 45, 20 and 30 seconds (125 seconds) into 2 billed minutes with 5 seconds carried. Calls marked no-answer or hangup_on_voicemail “are recorded as a flat 5 seconds of usage”; failed and user-canceled calls are not billed.
ElevenLabs’ cost help article says voice calls “are charged based on the call duration, with a 95% discount for periods of silence longer than 10 seconds”, measured as “the connection duration”, so “the call itself may be shorter than the duration you are charged for”. We could not find a stated rounding increment on its agents pricing page, that article or its Pay As You Go page, as at 6 November 2026; its changelog for 26 October 2026 says “teardown time after the call is no longer billed or counted”. Then Twilio rounds “any partial minutes (under 60 seconds)” up to the next full minute, which is why our 2 minute 30 second calls bill 3 carrier minutes each.
Rather not build the agent at all? Zian answers inbound calls 24/7 as an AI receptionist, books appointments and follows up by voice, SMS, email, WhatsApp, Instagram DMs and Facebook Messenger. It is in partnership-application beta.
Native numbers, SIP trunks and Twilio on Synthflow and ElevenLabs
| Telephony question | Synthflow | ElevenLabs Agents |
|---|---|---|
| Buy a number from the platform | Yes: US, Canadian and Australian numbers at US$1.50 a month “for quick testing” (Australian numbers need a registered address) | We could not find an option to buy a number for ElevenAgents in its docs index, as at 6 November 2026; ElevenLabs’ separate Reception.ai product assigns US and Canadian numbers |
| Your own Twilio account | Connect with Account SID and Auth Token; “Numbers stay in Twilio while routing, recording, and call transfers run through Synthflow” | Native integration: purchased Twilio numbers (inbound and outbound) or Verified Caller IDs (outbound only); or the register-call route, where call transfer “is not available” |
| SIP trunking | “SIP/PBX integration requires an Enterprise plan subscription”; Tier 1 native: Twilio, Telnyx, RingCentral RingCX, Vonage, 8×8, Zoom, Five9; Tier 2 (deployed with customers but not regression-tested): Dialpad, 3CX, Aircall, Genesys Cloud, Talkdesk | Inbound and outbound SIP trunking with TLS; guides for Telnyx, Plivo, Exotel, Bandwidth, Sinch and Vobiz, a Vonage WebSocket connector, plus Amazon Connect, Genesys, Five9 and Microsoft Teams |
| Regions | Global, US and EU regions on its own session border controllers; a closer point of presence on request | Static-IP SIP servers in the US, EU, India and Singapore for Enterprise accounts |
| WhatsApp Business calling to an inbound agent; WhatsApp and SMS chat need your own Twilio account and a US-region workspace | WhatsApp messages (text, voice notes, media) and inbound and outbound WhatsApp calls |
Synthflow’s telephony overview recommends a purchased number for testing and Twilio or SIP for production “so you keep control of your numbers, regional rules, and configuration”. Because every new Synthflow customer is on an Enterprise contract, its SIP gate mainly affects older self-serve workspaces.
No-code builder or developer platform?
Both are both, so decide by who will maintain the agent. Synthflow’s create-an-agent guide offers Prompt Builder (“one main prompt for simpler setups”) or Flow Designer (“nodes and branches for structured, multi-step flows”), templates, JSON import and an API, and its Enterprise docs describe a workflow builder that connects CRMs, scheduling tools and APIs “without writing code”. ElevenLabs’ agents page says you can build a first agent in minutes with “no engineering required”, its Workflows editor is a visual graph, and the same agent can be pulled and pushed as JSON through its CLI or SDKs, with WebSocket and REST APIs and native web and mobile SDKs.
The real split is the model and the support. On Synthflow, GPT models are an Enterprise feature, and its general configuration page says “The AI Model selector does not support Claude or custom LLM endpoints”; its configuration page names an OpenAI API key as the only bring-your-own option, and its integrations page also lists an Azure OpenAI route. ElevenLabs lists OpenAI, Anthropic and Google models plus a custom LLM option. Synthflow sells hands-on launch help into the contract; ElevenLabs lists Forward Deployed Engineers under ElevenAgents for Enterprise and leaves self-serve plans to self-serve.
How many calls at once each plan allows
| Capacity question | Synthflow | ElevenLabs Agents |
|---|---|---|
| Concurrent calls | “Scoped with Synthflow, shown in Plan & Billing” for Enterprise; legacy Pro, Growth and Agency workspaces keep fixed limits | Free 4, Starter 6, Creator 10, Pro 20, Scale 30, Business 40; Enterprise “Elevated concurrency limits” |
| Above the limit | HTTP 429 or a brief queue, depending on setup | Burst at US$0.160 a minute up to 3 times the limit or 300; call queueing holds callers, with “nothing is charged while the caller waits” |
| Bulk outbound | Campaigns, “available on our Enterprise plans”, 1 to 10,000 recipients per request | Batch calling from a CSV or XLS list; batches use the lower of 50% of the workspace limit or 70% of the agent’s |
So a 10,000-minute month on ElevenLabs Pro gives 20 simultaneous calls, but a batch campaign on it runs at most 10 at once. ElevenLabs’ burst guide also warns that burst calls get “lower priority for speech-to-text and text-to-speech processing”. On Synthflow, write your peak simultaneous calls and your calls-per-second need into the Order Form; its concurrency page also sets a separate calls-per-second (CAPS) limit, which warm transfers count against, and tells Enterprise customers to agree volume, queueing and limits with Synthflow before launch.
Languages: what each vendor’s own pages say
Synthflow’s languages page lists 33 distinct languages (Portuguese twice, Brazilian and European; Hebrew is printed twice in the list), including English with an American, British or Australian accent, plus a Multilingual mode that switches between 10 languages. ElevenLabs’ pages give three counts: its agents page says “90+ languages” and that agents “can switch between them mid-conversation”; its docs overview says 31 languages for configuration and 70+ for speech; its language guide says the All option gives 31 languages and that “Language selection is fixed for the duration of the call”, while its language detection tool switches language when a caller speaks another one. Test your languages on a live call before relying on any count.
HIPAA, SOC 2 and GDPR as each vendor states them
| Item | Synthflow | ElevenLabs Agents |
|---|---|---|
| Certifications and reports | Pricing FAQ: Trust Center lists “SOC 2 Type 2, ISO/IEC 27001, GDPR, HIPAA (self-attested) and PCI DSS v4.0.1”, with ISO/IEC 42001:2023, DORA and the EU AI Act “in progress”; its security page labels ISO/IEC 42001 and the EU AI Act “AI governance aligned” | Agents page: “SOC 2 Type II, ISO 27001, and PCI DSS Level 1 certified, with HIPAA and GDPR compliance”; ISO 42001 and AIUC-1 badges |
| HIPAA BAA | Signs a BAA with customers using its HIPAA Setting, which is off by default; BAA template published in its docs | BAA execution “is only available for Enterprise tier subscriptions”; Zero Retention Mode required, and it “cannot be enabled for batch calls” |
| Data location | Choose a US or EU data region per workspace | Data residency in the US, EU and India, listed among Enterprise controls |
| DPA | “DPA available” for GDPR; MSA and DPA support in Enterprise contracts | Enterprise plan: “Custom terms & assurance around DPA/SLAs” |
Read the status words, not the badges: Synthflow calls its own HIPAA position “self-attested” on its pricing page, while its security page heads the same item “BAA available”, and ElevenLabs’ HIPAA page says the BAA covers an agent only while Zero Retention Mode is on. For a HIPAA outbound programme on ElevenLabs that means placing calls one by one rather than in batches. The HIPAA BAA chain for AI voice agents explains which other parties, such as the carrier and the model provider, need their own agreement.
Integrations and channels each one connects
Synthflow’s integrations overview names GoHighLevel, Salesforce, HubSpot, Zendesk, Freshworks, Cal.com, WhatsApp, Zapier, Make, Bubble and ElevenLabs voices, and its changelog added ElevenLabs v4 voices on 7 October 2026. ElevenLabs’ docs list HubSpot, Salesforce, Zendesk, Freshdesk, Intercom, ServiceNow, Jira, Cal.com, Calendly, Google Calendar, Slack, Telegram and Twilio, and its agents page logo strip adds Zapier. Text is priced differently: ElevenLabs charges US$0.003 per message, while Synthflow counts five AI messages as one usage minute. So 1,000 SMS follow-ups of six AI messages are 1,200 Synthflow usage minutes at your contract rate, or US$18.00 in ElevenLabs message fees plus the model; Twilio’s messaging fees apply on both.
Synthflow vs ElevenLabs Agents: which one to choose
| If your situation is… | Lean towards | Why, from the figures above |
|---|---|---|
| You need a price this week without a sales call | ElevenLabs Agents | Published plans, US$0.080 extra minutes and per-model estimates; Synthflow has “no rate by the minute” |
| You want someone else to build and launch the agent inside the contract | Synthflow | Launch support by forward-deployed engineers, where the agreement includes it |
| Your Synthflow quote is under the ceiling for your volume | Synthflow | About US$0.105 a minute all-in with telephony, or US$0.082 if you keep Twilio on both (GPT-5 Mini) |
| Calls are short and numerous, and the two rates are close | Synthflow, on the meter | Seconds pooled across calls and a flat 5 seconds for unanswered calls; ElevenLabs bills connection duration. If you bring Twilio to either platform, Twilio still rounds each call up |
| You need Claude or your own model endpoint | ElevenLabs Agents | Anthropic models and a custom LLM option; Synthflow’s selector “does not support Claude or custom LLM endpoints” |
| You need HIPAA cover on bulk outbound | Get it in writing from either | ElevenLabs’ BAA needs Zero Retention Mode, which “cannot be enabled for batch calls”; we could not find Synthflow’s Campaigns or security docs saying whether its HIPAA Setting covers campaigns, as at 6 November 2026 |
| You need EU hosting | Either | Synthflow EU workspace region; ElevenLabs EU data residency on Enterprise |
| You need India data residency or Singapore SIP | ElevenLabs Agents | US, EU and India residency; static-IP SIP in the US, EU, India and Singapore |
| You would rather not build at all | Zian, or LeadsNow.ai done-for-you | Ready-made receptionist and outbound agents; managed, pay-per-result service |
Who each is wrong for: Synthflow is wrong for a team that wants to pilot on a credit card, because there is no free plan, trial or public rate. ElevenLabs Agents is wrong for a team that wants the vendor to own the build and the launch without an Enterprise contract. If you are also weighing Retell, our Retell AI vs ElevenLabs Agents cost comparison prices the same ElevenLabs plans against a per-second rival.
The sentence on each vendor’s pages that cuts against this page
Against our ElevenLabs totals: the pricing page calculator labels its model figures “Estimated costs in USD, covering ElevenAgents platform voice and LLM models. Excludes any charges from your own telephony provider”, and a tooltip says the displayed minute estimates are for “a custom LLM option rather than the LLM integrations we provide”. ElevenLabs’ Pay As You Go page adds that “If your balance hits $0 and Auto Top Up is off (or fails), your service will pause immediately”, and caps top-ups without two-factor authentication at US$100 a month on Creator, US$200 on Pro and US$600 on Scale. Its cost help article says you need Pay As You Go credits “before you can exceed your credit quota”, so the extra minutes in our worked bill (US$58.00, US$700.96 and US$3,700.96) are prepaid, and the 10,000 and 50,000-minute months need two-factor authentication switched on. Against our Synthflow framing: its minute is pooled per second, which no carrier on our ElevenLabs bill matches, and its Enterprise page answers “What’s the minimum commitment for Enterprise?” only with “Enterprise terms are defined in your agreement”, so a quote may bundle allowances or carryover that a per-minute comparison misses.
What this page does not price: Enterprise contracts, ElevenLabs’ Fast LLM tier (a “per-token price premium”, added 26 October 2026), Voice Library voice surcharges, Twilio recording and answering machine detection, and international calls. In Synthflow’s changelog, non-enterprise accounts have run only on Synthflow LLM since 25 August 2026, and Campaigns arrived for Enterprise on 18 August. Zian and LeadsNow.ai publish no per-minute price, so neither is in the cost tables.
Frequently asked questions
Is ElevenLabs Agents cheaper than Synthflow?
Nobody can say from public pages alone, because Synthflow publishes no per-minute rate. What you can say is where the line sits. At 10,000 connected US minutes a month on your own Twilio numbers, ElevenLabs Agents comes to US$1,050.32 on the Pro plan with GPT-5 Mini, about US$0.105 a minute all-in, as at 6 November 2026. A Synthflow quote that includes telephony has to land below that rate to cost less; if you keep Twilio on both platforms, the ceiling drops to about US$0.082 a minute.
Does Synthflow publish a price per minute?
No. As at 6 November 2026, Synthflow’s pricing page says it “is sold as an annual enterprise agreement” and has “no free plan or trial and no rate by the minute”. The only fixed figure in its docs is US$1.50 a month for a phone number bought for testing. Its ROI calculator models cost at 0.15 per minute but labels the result “Illustrative estimate only”.
What does ElevenLabs Agents charge per minute?
ElevenLabs lists US$0.080 for each call minute beyond a plan’s included minutes, on all six plans from Free to Business, with speech to text and text to speech included. Plans run from US$0 (15 minutes) to US$990 a month (12,375 minutes). The language model is billed on top at a per-model rate, and your phone carrier bills its own charges; ElevenLabs “adds no telephony fee”.
Can I use ElevenLabs voices on Synthflow?
Yes. Synthflow’s ElevenLabs page lists ElevenLabs Turbo v2 and v2.5, Flash v2 and v2.5, Multilingual v2, V3, v4 and v4 Turbo as voice models inside Synthflow, and you can connect your own ElevenLabs API key for custom voices. So choosing Synthflow does not mean giving up the ElevenLabs voice; it changes who you buy the minute from.
Which one signs a HIPAA BAA?
Both, on their top tier. ElevenLabs’ HIPAA page says BAA execution “is only available for Enterprise tier subscriptions” and requires Zero Retention Mode, which cannot be used for batch calls. Synthflow says on its security and compliance page that it signs a BAA with customers using its HIPAA Setting, and every new Synthflow deployment is an Enterprise contract. Synthflow’s pricing page describes its own HIPAA status as “self-attested”.
How many calls can run at once on each?
ElevenLabs publishes the numbers: 4, 6, 10, 20, 30 and 40 concurrent calls on Free, Starter, Creator, Pro, Scale and Business, with burst to three times the limit (capped at 300) at US$0.160 a minute. Synthflow scopes concurrency and its calls-per-second limit into each contract and shows them in the workspace’s Plan & Billing page; it prints no default for new customers.
Does ElevenLabs bill Twilio Media Streams on top?
ElevenLabs’ native and register-call Twilio guides do not say, although its custom-LLM Twilio guide relays call audio through Media Streams. Its register-call guide says the returned TwiML “connects the call to your agent via WebSocket”, and Twilio’s US voice pricing page lists Media Streams at US$0.0044 a minute. We put that line in our worked bill; at 10,000 talk minutes in 2 minute 30 second calls it is US$52.80. Check your first Twilio invoice and remove it if it does not appear.
Where every figure on this page comes from
| Figure | Who published it | Link | Date read |
|---|---|---|---|
| Annual enterprise agreement; “no free plan or trial and no rate by the minute”; every new deployment on Enterprise; contract scope; Trust Center list incl. “HIPAA (self-attested)”; launch support | Synthflow | synthflow.ai/pricing | 6 November 2026 |
| Sales-led pricing; Plan & Billing sections | Synthflow | docs.synthflow.ai/billing | 6 November 2026 |
| Per-second, pooled metering; 125-second example; flat 5 seconds for no-answer; chat at 5 messages per minute; dashboard cost by provider | Synthflow | docs.synthflow.ai/usage | 6 November 2026 |
| Minimum commitment “defined in your agreement”; Enterprise scope | Synthflow | docs.synthflow.ai/enterprise | 6 November 2026 |
| Concurrency and CAPS scoped; 429 or queue; legacy plans | Synthflow | docs.synthflow.ai/concurrency | 6 November 2026 |
| Three telephony options; SIP requires Enterprise; Tier 1 and Tier 2 providers; regions | Synthflow | docs.synthflow.ai/telephony | 6 November 2026 |
| US$1.50 a month numbers, US, Canada, Australia; Australian address | Synthflow | docs.synthflow.ai/buy-a-phone-number | 6 November 2026 |
| Twilio account connection; numbers stay in Twilio | Synthflow | docs.synthflow.ai/twilio | 6 November 2026 |
| 33 distinct languages; Australian accent; Multilingual mode across 10 languages | Synthflow | docs.synthflow.ai/about-supported-languages | 6 November 2026 |
| GPT models Enterprise only; no Claude or custom LLM endpoints; model list | Synthflow | docs.synthflow.ai/general-configuration | 6 November 2026 |
| Prompt Builder and Flow Designer; chat via Twilio; Twilio messaging fees | Synthflow | docs.synthflow.ai/create-an-agent | 6 November 2026 |
| Integrations list | Synthflow | docs.synthflow.ai/about-integrations | 6 November 2026 |
| Campaigns on Enterprise plans; 1 to 10,000 recipients per request | Synthflow | docs.synthflow.ai/batch-calling | 6 November 2026 |
| ElevenLabs voice models inside Synthflow | Synthflow | docs.synthflow.ai/elevenlabs | 6 November 2026 |
| Certifications; BAA with HIPAA Setting; US or EU data region; DPA available | Synthflow | synthflow.ai/security | 6 November 2026 |
| Fees non-refundable; usage volume cannot be decreased; unused volume expires | Synthflow | GTC – Direct Use (US) | 6 November 2026 |
| Billing in USD; taxes on invoices; VAT number | Synthflow | docs.synthflow.ai/manage-billing | 6 November 2026 |
| WhatsApp Business calling to an inbound agent | Synthflow | docs.synthflow.ai/deploy-your-agent | 6 November 2026 |
| Yearly minutes × 0.15 in the calculator script; “Illustrative estimate only” | Synthflow | synthflow.ai/roi-calculator | 6 November 2026 |
| Changelog: Synthflow LLM only on non-enterprise (25 August 2026); Campaigns (18 August); ElevenLabs v4 voices (7 October) | Synthflow | docs.synthflow.ai/changelog | 6 November 2026 |
| Plan prices US$0 to US$990; included minutes 15 to 12,375; concurrency 4 to 40; US$0.080 extra minute; US$0.160 burst; US$0.003 text; no telephony fee; LLM on top; per-model estimates (GPT-5 Mini US$0.002380986, GPT-5.5 US$0.044607915, range US$0.000476 to US$0.231506); yearly prices; estimate footnote and tooltip; cancellation; taxes excluded | ElevenLabs | elevenlabs.io/pricing/agents | 6 November 2026 |
| “90+ languages”; mid-conversation switching; certifications; US, EU and India residency; “no engineering required”; integration logos | ElevenLabs | elevenlabs.io/agents | 6 November 2026 |
| Connection duration; 95% silence discount; Pay As You Go credits needed to exceed the plan | ElevenLabs | How much does ElevenAgents cost? | 6 November 2026 |
| BAA Enterprise only; Zero Retention Mode | ElevenLabs | HIPAA | 6 November 2026 |
| Batch concurrency 50% / 70%; no Zero Retention Mode for batch calls | ElevenLabs | Batch calling | 6 November 2026 |
| Burst to 3 times or 300; 2 times the rate; deprioritised processing | ElevenLabs | Burst pricing | 6 November 2026 |
| Call queueing; nothing charged while waiting | ElevenLabs | Call queueing | 6 November 2026 |
| Purchased Twilio numbers and Verified Caller IDs | ElevenLabs | Twilio native integration | 6 November 2026 |
| TwiML connects “via WebSocket”; no transfer on this route | ElevenLabs | Register Twilio calls | 6 November 2026 |
| SIP trunking inbound and outbound; static IPs in US, EU, India, Singapore | ElevenLabs | SIP trunking | 6 November 2026 |
| 31 languages with All; language fixed for the call | ElevenLabs | Language | 6 November 2026 |
| Language detection tool switches language | ElevenLabs | Language detection | 6 November 2026 |
| 31 languages for configuration; 70+ languages for speech | ElevenLabs | ElevenAgents overview | 6 November 2026 |
| Visual Workflows editor; CLI and SDK | ElevenLabs | Workflows | 6 November 2026 |
| WhatsApp messages and calls | ElevenLabs | 6 November 2026 | |
| Service pauses at US$0; top-up caps without 2FA; top-ups non-refundable, expire after 12 months | ElevenLabs | Pay As You Go | 6 November 2026 |
| Teardown not billed; Fast LLM tier (26 October 2026) | ElevenLabs | ElevenLabs changelog | 6 November 2026 |
| Reception.ai assigns US and Canadian numbers | ElevenLabs | Reception.ai phone numbers | 6 November 2026 |
| Twilio audio relayed through Media Streams on the custom-LLM route | ElevenLabs | Custom LLM integration | 6 November 2026 |
| Integration guides: HubSpot, Salesforce, Zendesk, Freshdesk, Intercom, ServiceNow, Jira, Cal.com, Calendly, Google Calendar, Slack, Telegram, Twilio | ElevenLabs | ElevenLabs docs index | 6 November 2026 |
| US local outbound US$0.0140; local number US$1.15; Media Streams US$0.0044; recording US$0.0025; answering machine detection US$0.0075 | Twilio | twilio.com/en-us/voice/pricing/us | 6 November 2026 |
| Partial minutes rounded up on five named products incl. Media Streams | Twilio | Twilio support article 223132307 | 6 November 2026 |
| Worked bill totals, per-minute ceilings, yearly and GPT-5.5 variants, SMS comparison | Zian, calculated from the rows above with stated assumptions | This page | 6 November 2026 |
| 50,769+ qualified sales appointments set | Zian | zian.ai home page counter | 6 November 2026 |
Want the receptionist and the outbound follow-up on one platform? Zian’s AI Receptionist and Outbound Appointment Setter handle answering machines and voicemail on calls, work in 30+ languages and connect to HubSpot, Salesforce, HighLevel and Zapier. Each of its three listed plans, including Intel + Widgets (Free Forever), starts with a partnership application.