The software handles the mechanics. You handle the conversation.
DialBreeze is a browser power dialer with after-call AI. An operator works their own calling account, sees the lead context beside the call controls, and gets a usable record when the call ends.
That shaped a specific product: up to three lines for one operator, human-led connected calls, and a transcript, fixed structured fields and summary written after the hang-up. The calling list, the operator's work and the follow-through still determine what happens after a dial.
It is built for teams that sell by phone. It is not a source of leads or a promise of appointments.
The published throughput snapshot
Measured in production use. Aggregate, anonymized. Last 90 days, as of 2026-09-26. Separate activity counts, not a joined sales funnel or a forecast for your team. How we count.
Separately reported rates: last 30 days
- 17.8% person-connect
- Completed human-answer attempts divided by call attempts in the reported 30-day cohort.
- 26.7% voicemail
- Voicemail outcomes divided by dials in the reported 30-day cohort.
- 78% AI coverage of connects
- Human connects with a summary and opportunity flag divided by human connects in the reported 30-day cohort.
These owner-supplied rates are reported separately from the output totals. The published extract does not establish a one-to-one join between those measures. Do not apply these rates to the 90-day counts or reconstruct a funnel from them. Measured in production use, aggregate and anonymized, as of 2026-09-26.
Inventory at the snapshot, not period results
15,694 leads in the system all-time; 33 active lists as of 2026-09-26. These are inventory measures, not leads acquired or lists created in the last 90 days.
What these numbers measure.
The snapshot is dated September 26, 2026 and describes 90 days of production use, in aggregate and without identifying lead information. The 90-day activity counts cover dials, recordings, AI summaries and appointments set. The 30-day rates describe a separate, shorter reporting window. Lead and list totals describe the inventory in the system, not new leads or lists generated during either window.
A dial is an attempt, not a unique person reached. The same lead can receive more than one eligible attempt. A recording is an audio record, not proof that a qualified conversation occurred. An AI summary is a processing output, not an appointment. An appointment set is not an appointment held, a signed agreement or a closed transaction.
The person-connect rate is reported against attempts in its 30-day cohort. The voicemail rate is reported against dials in that cohort. AI coverage of connects uses human connects as its denominator, with a summary and opportunity flag as the reported qualifying output. These denominator labels matter more than a large number printed without context.
What the extract does not establish.
The published counts and rates are separate measures. The extract does not provide the matched call-level records needed to turn every recording and summary into a single conversion funnel. It also does not establish that all processing outputs were created on the same date as their underlying call. Applying the 30-day percentages to the 90-day counts would mix cohorts and produce an unsupported result.
For that reason, we do not infer an end-to-end conversion rate, average customer result or return on investment from this snapshot. It is evidence of throughput in production use. It is not a controlled comparison, a multi-customer benchmark or a forecast for another team with a different list, market, operator or calling policy.
The benchmark article owns the published dataset and its interpretation. The measurement guide explains how to define your own cohort and record a useful baseline without copying our results into your plan.
What the product views show.
The operator screenshots show captured interface states with test data and masked identifying details. They help you inspect where the list, lead context, call controls, outcomes and follow-up task sit. They are not customer testimonials or evidence of a live prospect's results.
A screenshot is not a substitute for inspecting a result. During evaluation, ask to review a permitted recording and its actual transcript, structured fields and summary together. Compare the source and output, including any names, dates or next-step suggestions that need correction. That is the useful test of the after-call workflow.
Why the human stays on the call.
The operator asks the questions, listens, handles uncertainty and agrees the next step. After-call AI can help turn a recording into something easier to review, but it does not decide what a person intended with certainty. It can miss words, confuse details or suggest a follow-up that needs correction.
We treat the transcript, structured fields and summary as different review aids. The transcript is an attempt to capture the words. The structured fields are signals. The summary is a compact handoff. None of them replaces the operator's responsibility to record an opt-out, confirm an appointment or check a material commitment against the audio.
How we charge, and who helps.
We sell a monthly software subscription: Starter is $25 for one dialing seat, Pro is $49 for one dialing seat with three lines and after-call AI, and Team is $79 for two dialing seats plus $20 for each added dialer. Only dialing seats are billed. Teams of 10 or more dialers scope Enterprise onboarding with us. Your own Telnyx account bills the carrier services separately; we do not resell those minutes or charge per dial.
Stripe processes billing. Checkout and Stripe receipts show the merchant name MIRA, DialBreeze's parent company. Your card statement shows DIALBREEZE. Contact support for subscription changes.
Accounts are provisioned and access instructions are emailed when they are ready. Same-business-day provisioning is a target rather than a guaranteed delivery time. We distinguish a received request, a successful payment and a ready account because they are different operating states, not interchangeable messages on a website.
Get started with assisted setup.
Setup is assisted. A person on our team creates your workspace, admin login and dialing seats after you choose a plan or send a setup request; there is no instant self-serve signup. Calls run on your own Telnyx account (bring your own), which Telnyx bills separately, and need an eligible calling list. The request form takes no payment and does not create an account.
We target same-business-day workspace setup once a plan is paid. This is a target, not a guaranteed access time. Your access and connection instructions arrive by email when the workspace is ready.
Tell us how your operator works day to day: list preparation, audio setup, the connected-call screen, a saved outcome and the after-call record. Bring your actual questions about importing data, handing work back to your CRM and supporting a teammate, and we will tell you plainly whether the workflow fits before you pay.