Synthflow vs ElevenLabs Agents: Contract or Monthly Plan

Synthflow suits phone operations that want launch help. It is sold as an annual enterprise agreement scoped to call volume and setup. ElevenLabs Agents suits builders who start today, with monthly plans and $0.08 a minute past included minutes. Dograh is the open-source AI voice agent platform we build, which you can self-host or run on Dograh Cloud.

Key Takeaways

• Synthflow is bought as one annual agreement, scoped to your volume and setup.
• ElevenLabs Agents runs on monthly plans, with the language model billed on top.
• Over the call limit, Synthflow returns a 429 or queues, ElevenLabs bursts at double rate.

Synthflow agents can speak with ElevenLabs voices, so picking between them is really about how you buy and run the agent. We compare them in one table, then on buying, launch, call limits, compliance and hosting, and who each fits. If you want open source, the last section covers Dograh.

Synthflow and ElevenLabs Agents in one table

The two vendors sell the same kind of product in very different ways, and the buying model shapes every other row. Dograh's column is there for the last section.

A scoped contract or a monthly plan

The biggest difference is how you pay, and it decides who can even start. Synthflow sells one plan, an annual enterprise agreement scoped to your call volume, phone system, integrations, security needs and launch support. There is no free plan, trial or per-minute rate. You talk to sales, agree the scope and get one price for the year.

ElevenLabs Agents works the other way round. ElevenLabs' agent pricing lists monthly plans from Free with 15 minutes up to Business at $990 a month with 12,375 minutes, with Starter, Creator, Pro and Scale in between. Past the included minutes, each extra minute costs $0.08. That minute covers ElevenLabs' own speech-to-text and voices, while the language model is billed on top at the model's rate and your phone carrier bills its own minutes.

So the two budgets look nothing alike. A Synthflow buyer plans one annual number with the vendor up front. An ElevenLabs buyer can start on a $6 plan this afternoon and watch the minute bill grow with traffic. For what the Synthflow agreement covers line by line, see what Synthflow's contract covers, and for every ElevenLabs plan, our breakdown of ElevenLabs pricing.

How each one gets an agent live

Synthflow sells help with the launch, and ElevenLabs sells tools you use yourself. Where the agreement includes it, Synthflow's forward-deployed engineers implement, test and launch the agents with your team, and Synthflow says agents go live in weeks. Synthflow lists tested integrations with Zoom Phone, Zoom Contact Center, Five9, RingCentral RingCX, Twilio, Telnyx, Vonage and 8x8, and calls can run on Synthflow's own telephony, a SIP trunk or approved enterprise telephony.

ElevenLabs Agents is built for a team that does the work itself. You get a visual workflow builder, automated tests, and SDKs for React, iOS, Android, React Native and Unity, so the same agent can sit in an app as well as on a phone line. Telephony is included on every plan in the sense that you connect your own provider, a native Twilio link or a SIP trunk, and the carrier bills you at cost.

The trade is time against control. A contact centre on Five9 or Zoom gets an agent wired into the system it already runs, without hiring voice AI engineers. A product team gets an agent today, but owns every integration and every test.

Busy hours and call limits

Both vendors cap how many calls run at once, and they handle the next call very differently. On Synthflow, concurrency is planned inside the contract. Synthflow's concurrency docs say that when you reach a limit, extra calls either return an HTTP 429 error or wait briefly in a queue, depending on your account setup.

Synthflow adds a second limit called CAPS, the number of outbound calls your account can start per second. Every transfer that dials the destination first starts a new outbound call, so it counts against CAPS too. Subaccounts share the main account's limits, and Synthflow advises a 1-second delay between calls when a CRM or Zapier triggers them.

ElevenLabs ties concurrency to the plan, from 4 calls at once on Free to 40 on Business. With burst pricing switched on, an agent can take up to three times its usual limit, and the extra calls cost double, $0.16 a minute. Above that, calls are rejected unless you turn on call queueing, which holds callers for 1 to 1,800 seconds (180 by default) and does not bill the waiting time.

For an outbound team, CAPS and burst matter as much as the price per minute. Our guide to voice agent platforms for high call volume sets every vendor's limit side by side.

Open Source Alternative to Vapi / Retell

Self-hosted voice agent platform — no per-minute fees

dograh-hq/dograh

Star on GitHub

Compliance, hosting and agency use

Synthflow lists more certificates, and ElevenLabs puts its HIPAA agreement behind Enterprise. Synthflow's Trust Center lists SOC 2 Type 2, ISO/IEC 27001, GDPR, PCI DSS v4.0.1 and HIPAA, which Synthflow marks as self-attested, with ISO 42001, DORA and the EU AI Act in progress.

ElevenLabs offers a business associate agreement (BAA) under the Health Insurance Portability and Accountability Act (HIPAA) on its Enterprise plan. Private deployment, where the agent runtime runs in your own AWS or GCP account, is also arranged through ElevenLabs' sales team. Otherwise both run your agent and keep your call data on their own cloud.

Agencies and BPOs have one more row to check. Synthflow lets them run it under their own brand inside the same agreement, with a workspace per client and one analytics dashboard across them. That makes the annual agreement easier to resell, since the clients never see Synthflow's name.

Which team each one suits

Synthflow suits a phone operation that already runs a contact-centre system and wants an agent inside it without building a voice team. That buyer pays for engineers who help with the launch and for tested links into Five9 or Zoom. The trade is no trial and a sales cycle before the first call.

ElevenLabs Agents suits a builder who wants to start today and cares most about how the agent sounds. Plans start small, and the SDKs reach apps as well as phones. The trade is that a HIPAA agreement and more than 40 calls at once both need an Enterprise deal, and the language model and carrier stay separate bills.

Before you commit to either, put the agent on a real phone line. Most phone calls still reach the agent as 8 kHz audio, while a browser demo runs at 16 kHz, so a voice that sounds perfect in a demo can mishear callers on the phone.

Dograh, the open-source third option

If you want ElevenLabs' voices or a Synthflow-style call flow without either vendor's contract, Dograh runs the agent on your terms. The code is open source under a BSD 2-Clause licence, and you pick the models. ElevenLabs voices run inside Dograh with your own ElevenLabs key, and so does any other language model, speech-to-text or voice provider, or open models on your own servers.

Self-hosted Dograh has no per-minute platform fee, but servers, model keys and phone minutes still cost money. Dograh Cloud charges a 1¢-a-minute platform fee. You can self-host it with Docker or start on Cloud the same day, with no annual agreement either way.

Call limits are settings you control. A self-hosted install runs 10 calls at once by default, and you raise it with one setting, while servers and your carrier set the practical ceiling. Each campaign also has its own maximum concurrent calls and calls started per second, which starts at 1, so you pace outbound dialling yourself. An agency can run its own copy of self-hosted Dograh on its own domain, under its own name and logo, as long as it keeps the licence notice.

The voice can be part recording. Pre-recorded clips play the lines your agent says most, and a clone of the same voice speaks the rest, so the whole call keeps one voice and the common lines start without waiting for text-to-speech.

Quality checks come built in. Dograh's QA scores each step of test and live calls with tags, a 1 to 10 score, sentiment and a summary, on the share of calls you choose. Start with test calls before the first live traffic.

When a score looks wrong, the call trace shows the speech-to-text, AI and voice time for every turn, so you can see which part of the call slowed down.

For the wider field, see our guides to Synthflow alternatives and ElevenLabs alternatives.