TL;DR. Bitrix24 has its own AI — BitrixGPT, called CoPilot in English-language Bitrix24. It transcribes a call, writes a summary, fills the CRM card and scores the conversation against a sales script you write, with a rating per manager. For a team that wants a script check inside the portal, that can be enough. Aelo differs in three ways. It analyses every call automatically, where BitrixGPT runs each one by hand. Its score is built from structured criteria, with quotes from the conversation as evidence. And it costs one fixed rate per analysed minute, with customer content stored at rest in the EU (in the database and the audio storage; the search index — numerical vectors, no text — and Cloudflare’s processing state, queues, logs, caches and coordination state, some of which hold transcript and chat text for a few days, are not pinned to the EU). It isn’t either/or — it’s a question of whether you need systematic quality control or a script check in the box is enough.
Quick verdict
| Dimension | Aelo | Bitrix24 (built-in AI) |
|---|---|---|
| Call coverage | Every call, automatically, once Automatic mode is on (a new organization starts in Manual); calls with under 30 s of conversation are skipped by default — a project setting | Each call is started by hand |
| Scoring criteria | 25 parameters, per-call-type profiles, script steps and required phrases, words that must not be said, your own criteria — findings backed by quotes | Sales scripts in free text (one or several per call): a match score, notes on what matched and what didn’t, a rating per manager |
| CRM card | What you set up: a report in the timeline, the empty fields you select filled in, dated commitments as planned calls or to-dos — each its own setting | Fills empty fields and proposes changes to filled ones (string, list, number), creates an activity with the outcome, flags spam |
| Economics | Transparent: one fixed rate per analysed minute, cost of every analysis in the billing log | Each run draws on the portal’s AI request allowance |
| Sources | Any telephony that exposes recordings, plus chats; the public API (Starter and above) for recordings that never reach Bitrix24 | Calls with a recording in the CRM timeline, plus chats, emails and video calls |
| For integrators | White-label, multi-tenancy | None |
| Data | Customer content stored at rest in the EU (in the database and the audio storage; the search index — numerical vectors, no text — and Cloudflare’s processing state, queues, logs, caches and coordination state, some of which hold transcript and chat text for a few days, are not pinned to the EU); public DPA and sub-processor register | Depends on your portal’s region — verify with Bitrix24 |
| Getting started | Self-serve, 100 free analyzed minutes monthly, no card | Within your existing Bitrix24 plan |
Independent comparison, checked against Bitrix24 documentation on 2 October 2026. Bitrix24’s built-in AI is evolving quickly — confirm current capabilities in Bitrix24 documentation before deciding.
What Bitrix24’s built-in AI does
To be fair: BitrixGPT is a competent AI in the box. In the CRM it transcribes a call, writes a summary, fills empty fields on the lead or deal card — and proposes a change when a field already has a value — and creates an activity with the outcome, or flags the call as spam (Bitrix24: CoPilot in CRM). Its speech analytics scores a call against a sales script you write in plain text — the same call can be scored against several scripts — points out what matched and what didn’t, and keeps an overall rating per manager (Bitrix24: AI Speech analytics and Sales scripts); the scores are also available as BI Builder datasets. Outside the CRM, CoPilot Follow-Up summarises video calls (Bitrix24: CoPilot Follow-Up). If a script check inside the portal is what you need, the built-in AI is a reasonable choice without a separate product.
Where the built-in AI hits its ceiling
- A manual run per call. Each call is processed by hand and draws on the portal’s AI request allowance (Bitrix24, in Russian). Fine for spot checks; not for blanket quality control of every conversation.
- A script, not a scorecard. A call is scored by how closely it follows a script written in free text — one score per script, with notes on what matched. There is no structured scorecard behind it: separate parameters with their own scores, a list of words that must not be said.
- The summary stays around the script. The manager rating and the BI datasets aggregate script scores — useful, but they answer one question: how closely the script was followed.
- Only calls in the portal. BitrixGPT analyses a call recorded in the CRM card. A recording that never reaches Bitrix24 isn’t analysed.
- No white-label or multi-tenancy — critical for integrators running many client portals.
- Data residency needs verifying. Bitrix24 runs cloud portals in several regions, so where your recordings are stored and where its AI processes them depends on your portal’s region and plan — check it against Bitrix24’s own documentation before you rely on it for compliance. Aelo publishes a single answer for every customer: content stored at rest in the EU (in the database and the audio storage; the search index — numerical vectors, no text — and Cloudflare’s processing state, queues, logs, caches and coordination state, some of which hold transcript and chat text for a few days, are not pinned to the EU), Cloudflare edge processing may occur outside the EU, with a published DPA and sub-processor register.
What Aelo does differently
- Automatic analysis of every call. Once you turn on Automatic mode, every conversation is analysed with no manual run and no BitrixGPT subscription, so quality control stops depending on who had time. Calls with under 30 seconds of conversation are skipped by default; the threshold is a project setting.
- Your own scoring criteria. Prompt Builder: 25 parameters across 8 groups and per-call-type profiles. Your script’s own steps and required phrases live in Script rules, and the words that must not be said are a per-project list under Analysis & Quality — all without a developer.
- Voice of Customer across the flow. Besides scoring each call, Aelo adds up what customers say over a period: objections by reason, competitors and how they were spoken of, product feedback, market signals and needs (Voice of Customer, Pro and above).
- Transparent economics. You are billed per analysed minute at your plan’s fixed rate (chats per chat), and the cost of every analysis appears in the billing log. Billing is prepaid: processing runs only while there is balance, so no invoice can outrun what you put in.
- Omnichannel. Any telephony that exposes recordings, plus chats; 90+ languages.
- White-label and multi-tenancy for integrators: per-client data isolation, one panel, per-client cost report.
- EU primary storage for recordings and transcripts. Customer content is stored at rest in the EU (in the database and the audio storage; the search index — numerical vectors, no text — and Cloudflare’s processing state, queues, logs, caches and coordination state, some of which hold transcript and chat text for a few days, are not pinned to the EU); Cloudflare edge processing may occur outside the EU. The data processor is Auspex Streamline S.L. (Spain), with a published DPA.
- You need findings you can audit. Aelo’s findings come with the quote from the transcript they rest on — evidence you can open, not an opinion — and every verdict traces back to the scorecard version behind it.
Aelo runs natively on top of Bitrix24 — an embedded app from the Bitrix24 Marketplace — and isn’t tied to the built-in AI: it’s a standalone product, not an add-on to BitrixGPT.
Bitrix24 call analytics via API: build or buy
Bitrix24’s REST API hands you raw call data, not analytics: voximplant.statistic.get
returns call history, telephony.externalCall.* registers calls, and transcription,
scoring and reporting stay yours to build and maintain. Aelo is that missing
analytics layer — it writes finished scores onto Bitrix24 records natively, and if
you’d rather integrate yourself, Aelo’s public call analytics API
returns the same structured analysis (summary, quality score, per-criterion
metrics) as JSON.
Pricing comparison
Aelo is priced on one axis — analyzed minutes — and its tiers are public (pricing page):
- Pay as you go — free. 100 analyzed minutes a month, then 6.5¢ per minute.
- Starter — $59/mo. 1,000 minutes (5.9¢/min), plus the public upload API.
- Pro — $129/mo. 2,500 minutes (5.2¢/min).
- Business — $269/mo. 6,000 minutes (4.5¢/min).
- Scale — $559/mo. 14,000 minutes (4.0¢/min).
- Enterprise. Custom volume and rate (≤3.5¢), SLA. Annual billing saves 20%.
Bitrix24’s built-in AI ships with Bitrix24, and its call processing in the CRM is available on commercial plans (Bitrix24, in Russian). Each run draws on the portal’s AI request allowance, which depends on the plan. It doesn’t expose the kind of per-call cost breakdown Aelo does — check current Bitrix24 terms.
Frequently asked questions
Bitrix24 already analyses calls — why Aelo?
BitrixGPT scores calls against your sales script inside the portal, and each call is started by hand. You need Aelo when you build quality control systematically: every call analysed automatically, structured criteria backed by quotes from the conversation, transparent economics, and EU primary storage for recordings and transcripts.
Does Aelo analyse every call automatically?
Yes, once Automatic mode is on. A new organization starts in Manual mode, so you can set up criteria before minutes are spent; one click on the Records banner or in organization settings switches it to Automatic. From then on every call with a conversation is analysed; calls with under 30 seconds of conversation are skipped by default, and the threshold is a project setting. The verdict reaches the Bitrix24 record the way you set it up: a report with the score and flags in the timeline; the empty fields you select filled in; commitments with a date as planned calls or to-dos. Each of these is its own setting. You can switch back to Manual at any time, and an organization on a paid plan that runs out of balance is paused into Manual — topping up does not restart analysis by itself.
Can I configure my own scoring criteria?
Yes. In Prompt Builder you enable the parameters you need (25 across 8 groups) and pick per-call-type profiles; your script’s steps and required phrases go into Script rules, and the words that must not be said into the project’s Analysis & Quality settings. The metric once labelled “Stop words” is now Agent uncertainty — it measures hedging about price, deadlines and promises from a fixed list, and is not the editable one.
Do I need a BitrixGPT subscription to use Aelo?
No. Aelo is a standalone product installed from the Bitrix24 Marketplace and doesn’t require the built-in AI subscription.
Does Aelo work with non-Bitrix24 telephony?
Yes. Aelo ingests recordings from any telephony that exposes them, not only Bitrix24 telephony — including recordings that never reach the Bitrix24 portal, through the public API, available on Starter and above.
Can I get Bitrix24 call analytics through an API?
Yes, two ways. The native integration writes scores onto Bitrix24 call records
with no code; Aelo’s public API lets you upload call recordings
by file or URL and read the structured analysis back. Bitrix24’s own REST methods
(voximplant.statistic.get, telephony.externalCall.*) expose raw call data but
no scoring — that part you’d have to build yourself.
Where is data stored?
Customer content is stored at rest in the EU (in the database and the audio storage; the search index — numerical vectors, no text — and Cloudflare’s processing state, queues, logs, caches and coordination state, some of which hold transcript and chat text for a few days, are not pinned to the EU); Cloudflare edge processing may occur outside the EU. The data processor is Auspex Streamline S.L. (Spain). The DPA is published at /en/legal/dpa/.
Where can I see live pricing?
Aelo pricing is at /en/pricing/. For Bitrix24’s built-in AI terms, check current Bitrix24 documentation.