Synthflow AI pricing

About

Synthflow AI is a no-code platform for building AI voice agents that automate phone conversations at scale — answering, routing, qualifying, and booking across customer service, receptionist, and contact-center use cases. Founded in 2023 by brothers Albert and Hakob Astabatsyan together with Sassun Mirzakhan-Saky, the company positions itself as an enterprise AI voice platform and says it powers more than 65 million voice calls per month across 30+ countries, with 1,000+ G2 reviews.

In June 2025 Synthflow raised a $20M Series A led by Accel (with Atlantic Labs and Singular), bringing total funding to roughly $30M, and reportedly reached around $1.1M revenue with a roughly 10-person team. That enterprise repositioning is now fully reflected in pricing: as of June 2026 the public pricing page shows a single, contact-sales Enterprise plan with contracts starting at $30,000/year — the self-serve per-minute Pay-As-You-Go path Synthflow ran earlier in 2026 has been removed for new customers. The company first retired its cheap flat subscription tiers, then briefly published a usage-based per-minute model, and has now consolidated entirely on quoted enterprise deals.

For the most current information, visit Synthflow.

Pricing summary : How Synthflow’s pricing model works

As of June 2026, Synthflow’s public pricing page shows a single plan: Enterprise. There is no longer a self-serve, published per-minute path on the page. The card states that Enterprise contracts start at $30,000 annually, with final pricing scoped around call volume, concurrency, telephony setup, integrations, security needs, and launch support — every number beyond the $30,000 floor is quoted via Contact Sales.

The Enterprise plan, as described on the page, includes an Enterprise SLA and support terms scoped in contract, Synthflow Native Telephony / SIP trunking or approved enterprise telephony, custom concurrency planning, routing, escalation paths, handoffs and fallback logic, CRM / calendar / contact-center / webhook / API / knowledge-source integrations, MSA/DPA support with data-handling review, workspace controls and an enterprise security review, and implementation, onboarding, testing, training, launch support and ongoing optimization.

What makes this different (and what changed): earlier in 2026 Synthflow ran a no-platform-fee, composable per-minute model — billing the Voice Engine, the LLM, and telephony as three separate per-minute meters with several concurrent calls included. That entire self-serve path has been removed from the pricing page; Synthflow has consolidated on a quoted, sales-led Enterprise motion. The historical per-minute and flat-tier figures are preserved in Pricing evolution and Hidden costs; they are grandfathered for existing customers but are no longer offered or published for new signups.

Pricing by product

The current pricing page (captured 2026-06-24) lists one plan:

No longer offered to new customers (grandfathered for existing):

• Pay As You Go — the no-platform-fee, composable per-minute model (Voice Engine + LLM + telephony, with reserved-concurrency and white-label add-ons) that Synthflow ran until mid-2026. The exact per-minute figures are listed in Pricing evolution and Hidden costs.
• Legacy flat tiers — Starter, Pro, Growth, and Agency monthly subscriptions with bundled minutes and low-teens-of-cents overage (see Pricing evolution for figures).

Sales motions across products: Synthflow is now sales-led only on its public pricing surface — the single Enterprise plan routes entirely through Contact Sales, with no self-serve sign-up path published. The prior PLG / self-serve Pay-As-You-Go motion has been retired for new customers.

Hidden costs : What Synthflow users actually pay

The “$0.09/min” headline is only the voice engine — the real per-minute cost is a stack. Add the LLM and telephony meters and a typical agent runs $0.11–$0.24/min before any add-ons. The pricing page’s own calculator illustrates the build-up: Voice Engine $0.09 + GPT-4.1 $0.05 + native telephony = ~$0.14/min in its default example.

Other things to budget for: concurrency beyond the 5 included calls is $20/month per reserved slot (up to 50), phone numbers are $1.50 each, white-label/reseller is $2,000/month on PAYG, and HIPAA, 99.99% SLA, and native telephony are Enterprise-only. Because there is no included-minute floor on PAYG, light usage is cheap but high-volume callers are steered toward an Enterprise annual commitment.

Want to estimate your own Synthflow bill? Use the Synthflow pricing calculator to model your costs based on usage patterns.

Pricing evolution : Synthflow pricing history and changes

Cadence

Tracked range: 2024–present. Live captures 2026-06-09 (PAYG + Enterprise) and 2026-06-24 (Enterprise-only); legacy tier detail corroborated by third-party pricing write-ups (Wayback access was blocked at capture time, so historical snapshot metadata is left empty rather than guessed).

Notable changes

• 2024 — Sells flat subscription tiers: Starter $29/mo (50 min), Pro $375/mo (2,000 min, $0.13 overage), Growth $750/mo (4,000 min, $0.12 overage, 50 concurrency, 25 subaccounts), Agency $1,250/mo (6,000 min, $0.12 overage). 14-day free trial; $1.50/mo per phone number.
• 2025-06 — Raises $20M Series A led by Accel (~$30M total); pivots messaging toward enterprise-grade voice agents.
• 2026 H1 — Replaces flat tiers with Pay As You Go (no platform fee — $0.09/min Voice Engine + LLM $0.02–$0.05/min + telephony $0.00–$0.02/min; 5 concurrency included, then $20/mo each) and custom Enterprise (10K+ min/mo). Legacy Starter/Pro/Growth/Agency grandfathered for existing customers.
• 2026-06 (mid) — Removes the self-serve Pay-As-You-Go path from the pricing page entirely. The public page now shows a single Enterprise plan with contracts starting at $30,000 annually and a Contact Sales motion; final price is scoped per deployment. PAYG and the legacy flat tiers are grandfathered for existing customers only. This consolidates Synthflow on a fully sales-led, quoted model.

The Enterprise-only consolidation in detail

The 2026-06-24 change is the third and most decisive step in a two-year march upmarket, and it reverses Synthflow’s most recent transparency move. Confirmed against the prior 2026-06-09 capture, the page went from two paths (a public, self-serve per-minute PAYG model plus a custom Enterprise tier) to a single Enterprise card stating only that “Enterprise contracts start at $30,000 annually,” with call volume, concurrency, telephony, integrations, security, and launch support all scoped through Contact Sales.

Three things changed at once for a prospective buyer:

• Buyer impact: there is no self-serve path left. Builders, pilots, and SMBs who could previously ship an agent for $0-to-start and pay per minute now have to book a sales call and clear a $30,000 annual floor — the no-code product’s bottom-up funnel is closed for new signups.
• Cost impact: the only disclosed figure is the $30K/year floor. The composable Voice Engine + LLM + telephony meter (and its ~$0.11–$0.24 all-in band) still almost certainly underpins the quote, but buyers can no longer compute their own cost — everything above the floor is negotiated.
• Transparency impact: this is a public → gated shift. Synthflow swapped a live per-minute calculator for a single anchor number, trading self-serve clarity for sales-led qualification.

The move is consistent with a lean ~10-person, post-Series-A team concentrating GTM on large annual contracts (its open roles skew to enterprise AEs and a Forward Deployed Engineer). It is a deliberate positioning choice, not a price hike — existing PAYG and legacy-tier customers keep their plans and pricing unchanged.

What’s unique : Synthflow’s distinctive pricing mechanics

1. A pricing page that walked all the way up the ladder. Synthflow’s most distinctive move is the direction and speed of its packaging changes. In under two years it went from cheap flat tiers ($29 Starter), to a no-platform-fee composable per-minute model, to — as of 2026-06-24 — a single Enterprise plan with a $30,000/year floor and no published rate beyond it. Almost every voice-AI peer is moving the other way (publishing per-minute rates to win self-serve developers); Synthflow deliberately closed that door.

2. Disclose only the floor, quote everything above it. The current page is a textbook gated transparency model: it names exactly one number — “$30,000 annually” — and explicitly scopes everything else (call volume, concurrency, telephony, integrations, security, launch support) to a Contact Sales conversation. The single anchor price does real work: it pre-qualifies leads (small buyers self-select out) while keeping the actual deal price negotiable per deployment.

3. The composable per-minute meter is now a private, not public, mechanic. Earlier in 2026 Synthflow’s signature was billing the Voice Engine, the LLM, and telephony as three separate per-minute lines — buyers could swap models or bring their own Twilio ($0.00 telephony) to tune the rate. That unbundled meter still defines the product technically (it survives for grandfathered customers and almost certainly underpins the Enterprise quote), but it is no longer a published, self-serve lever — it has moved behind the sales motion.

Strengths & weaknesses

Billing UX : Synthflow billing controls and transparency

• Contract & quote motion — The public pricing page now exposes a single Contact Sales path: there is no self-serve sign-up or published rate card. Enterprise contracts start at $30,000/year and are invoiced under a custom annual agreement; final pricing is scoped around call volume, concurrency, telephony setup, integrations, security needs, and launch support.
• Contractual controls — Enterprise terms are negotiated in the contract: MSA/DPA support, a data-handling review, workspace controls, and an enterprise security review, plus a scoped Enterprise SLA and support terms (the prior public PAYG page advertised a 99.99% uptime SLA at the Enterprise level).
• Onboarding & scoping — Pricing is set during a sales/scoping conversation that covers implementation, onboarding, testing, training, launch support, and ongoing optimization, with custom concurrency planning, routing, escalation paths, handoffs, and fallback logic rather than self-serve add-ons.
• Historical (grandfathered) self-serve controls — The earlier Pay-As-You-Go page was fully self-serve with no platform fee, reserved-concurrency add-ons, per-number telephony charges, and an interactive cost calculator that showed a live Voice Engine + LLM + telephony per-minute breakdown (figures in Pricing evolution and Hidden costs). That self-serve surface has been removed from the pricing page; existing PAYG/legacy-tier customers keep their plan and price.

Strategic wins : Why Synthflow’s pricing decisions worked

1. Committing fully to a single enterprise motion

Rather than straddle PLG and sales — which often means two pricing pages, two support models, and a confused funnel — Synthflow collapsed everything into one quoted Enterprise plan on 2026-06-24. For a ~10-person team post-Series A, concentrating GTM on a handful of large, annual-commitment deals is far more capital-efficient than servicing a long tail of self-serve accounts. The hiring mix (enterprise AEs, a Forward Deployed Engineer) is now aligned with that single motion. See how AI companies structure pricing.

2. Publishing a floor instead of going fully dark

Synthflow could have hidden all pricing behind “Contact Sales” with no number at all. Instead it discloses a $30,000/year floor — a smart middle path that does the qualifying work for the sales team: budget-constrained buyers self-select out before booking a call, while the actual deal price stays negotiable per deployment. A visible anchor also keeps the page eligible for price-aware AI-search citations that a pure sales-only page forfeits. Related: outcome-based pricing trends.

3. Steering all value toward annual contracts

The progression — retire cheap flat tiers, then retire the no-fee per-minute path — routes every new customer into a quoted, annual-commitment Enterprise deal, capturing more predictable revenue from the accounts that matter most after the Accel round. The composable voice + LLM + telephony meter still underpins the quote; it has simply moved from a public lever to a private one. See choosing the right usage metric.

Areas to improve : Gaps in Synthflow’s pricing approach

1. Abandoning the self-serve funnel for a no-code product

Synthflow’s whole pitch is “no-code voice agents” — a category whose buyers expect to swipe a card and ship. Removing every self-serve path on 2026-06-24 in favor of a $30K-floor sales motion cuts off the SMB and builder funnel that fed the product’s bottom-up adoption. A concrete fix: keep a capped, published self-serve PAYG tier (even a low monthly minute ceiling) alongside Enterprise, so the product still has a top-of-funnel while large deals route to sales. See how AI companies are shifting pricing.

2. Going from a visible rate card to one disclosed number

Earlier in 2026 Synthflow was unusually transparent — a live calculator showing the Voice Engine + LLM + telephony build-up. The current page discloses only the $30,000 floor, so prospects can no longer estimate their own cost. A fix that preserves the qualifying benefit without the opacity: publish indicative per-minute Enterprise rate bands (e.g. “from $0.X/min at volume”) next to the annual floor, so buyers can sanity-check the deal before booking a call. See bill shock and cost unpredictability.

3. The $30K floor is a hard wall, not a ramp

Jumping from “$0 to start” to a $30,000 annual minimum leaves no middle step for the mid-market buyer who is too big for a pilot but not ready for a six-figure commitment. A scoped “growth” contract tier — say a lower annual floor with a usage cap — would give that segment a published on-ramp instead of forcing a binary self-select between grandfathered PAYG (closed) and full Enterprise. See choosing the right usage metric.

Key takeaways

1. You can move pricing upmarket by removing tiers, not adding them. Synthflow killed its $29 Starter, then its no-fee per-minute path, ending on a single $30K/year Enterprise plan within two years of its Series A.
2. Packaging can move fast in both directions. The same team that added a transparent self-serve PAYG model in early 2026 removed it months later — a reminder that a published price is a positioning choice, not a permanent commitment.
3. A disclosed floor beats a blank “Contact Sales.” Naming one number ($30,000/year) pre-qualifies leads and keeps the page citable in price-aware AI search, while everything above the floor stays negotiable.
4. Going sales-led has a funnel cost. Concentrating on quoted Enterprise deals suits a lean post-Series-A team, but it forfeits the bottom-up adoption a no-code product usually relies on — a trade other founders should weigh before pulling self-serve.
5. The meter can outlive the rate card. Synthflow’s composable voice/LLM/telephony metering still defines the product technically; it simply moved from a public, self-serve lever to a private input to the Enterprise quote.

UBP implications

1. Per-minute (media-minutes) remains the natural value metric for voice AI — it maps directly to what the buyer’s business produces (handled calls) and scales cleanly with usage, which is why Synthflow’s meter persists even after the public rate card came down.
2. A usage meter and a public price are separable decisions. Synthflow still bills voice + LLM + telephony per minute, but as of 2026-06-24 that meter sits inside a quoted annual contract rather than a self-serve rate card — usage pricing can run entirely behind a sales motion. See usage-based pricing strategy.
3. Annual contracts with a disclosed floor are a distinct transparency tier. Synthflow’s “$30,000/year, everything else quoted” sits between fully public usage pricing and a blank sales-only page — a viable model when the buyer is enterprise, but one that sacrifices the self-serve UBP funnel a no-code product would otherwise capture.

Sources

• Synthflow pricing page (live capture, accessed 2026-06-24; prior PAYG state captured 2026-06-09)
• Synthflow AI raises $20M Series A (Accel) (accessed 2026-06-09)
• Synthflow AI raises $20M — Yahoo Finance (accessed 2026-06-09)
• Decoding Synthflow Pricing and Plans in 2025 — Dograh (legacy tier detail, accessed 2026-06-09)
• Synthflow Pricing 2026 — PxlPeak (accessed 2026-06-09)
• Synthflow pricing in 2026 — Ringly (accessed 2026-06-09)
• Synthflow billing overview — docs (independent price-validation second source; confirms sales-led, scoped Enterprise pricing, accessed 2026-06-24)
• How Synthflow AI hit $1.1M revenue — GetLatka (accessed 2026-06-09)

Bottom line

Synthflow AI is a no-code AI voice-agent builder (founded 2023, $20M Series A led by Accel in June 2025) that has steadily moved its pricing upmarket. As of 2026-06-24 the public pricing page shows a single Enterprise plan — contracts start at $30,000/year, with everything above that floor scoped per deployment (call volume, concurrency, telephony, integrations, security) through Contact Sales. That completes a fast progression: cheap flat tiers (Starter $29 → Agency $1,250), then a transparent no-platform-fee per-minute model (Voice Engine $0.09 + LLM $0.02–$0.05 + telephony $0.00–$0.02, ~$0.11–$0.24 all-in), now removed for new customers in favor of a fully sales-led motion. The per-minute meter still underpins the Enterprise quote, and PAYG plus the legacy flat tiers are grandfathered for existing customers only. Browse the pricing blueprint for more fully-researched company profiles.

Want to compare Synthflow against other voice and conversational-AI companies? Browse the pricing blueprint.

Pricing timeline : Major events on a vertical axis

Each milestone below corresponds to a public pricing change, product launch, or material adjustment. Major events use a filled marker; minor adjustments use a faded one.

PAYG removed — pricing page is now Enterprise-only

Jun 2026

Synthflow removed the self-serve Pay-As-You-Go path from its pricing page. The public page now shows a single Enterprise plan: contracts start at $30,000 annually, with final pricing scoped per deployment (call volume, concurrency, telephony, integrations, security) via Contact Sales. The prior per-minute PAYG model and legacy flat tiers are grandfathered for existing customers only; the motion is now fully sales-led.

Pay-as-you-go usage model replaces flat tiers

Jun 2026

Pricing page shows two paths: Pay As You Go (no platform fee — $0.09/min Voice Engine + LLM $0.02–$0.05/min + telephony $0.00–$0.02/min ≈ $0.11–$0.24 all-in; 5 concurrent calls included, then $20/mo per reserved concurrency) and custom Enterprise (10,000+ min/mo, 99.99% SLA, native telephony/SIP, unlimited concurrency, white-label, HIPAA). Legacy Starter/Pro/Growth/Agency tiers grandfathered for existing customers.

$20M Series A (Accel) — enterprise repositioning

Jun 2025

Raised a $20M Series A led by Accel (with Atlantic Labs and Singular), ~$30M total funding, and repositioned around enterprise-grade AI voice agents. This precedes the shift away from low-end flat tiers toward usage-based and enterprise pricing.

Flat monthly tiers — Starter/Pro/Growth/Agency

Jun 2024

Synthflow sells flat subscription tiers with bundled minutes: Starter $29/mo (50 min), Pro $375/mo (2,000 min, $0.13/min overage), Growth $750/mo (4,000 min, $0.12 overage, 50 concurrency, 25 subaccounts), Agency $1,250/mo (6,000 min, $0.12 overage, 80 concurrency), plus custom Enterprise. 14-day free trial; $1.50/mo per phone number.