HexPay (also appearing under the name CallerAPI in official materials) is a caller identification, anti-spam, and anti-fraud data service for communications and voice platforms, with the core positioning of enabling platforms to both reduce the costs caused by spam calls and resell protection capabilities as a value-added service for profit.
[Purpose and Positioning] Materials show that its customer base includes MVNOs (mobile virtual network operators), CPaaS, UCaaS, and voice AI platforms. For MVNOs, CPaaS, and UCaaS, it can resell spam protection and caller intelligence as value-added services under their own brand, creating new revenue streams and reducing user churn; for voice AI platforms, it can perform call screening before agents or models answer, preventing fraudulent calls from entering the model and avoiding consumption of large-model tokens and telephony minutes by non-target calls.
[Main Capabilities] Officially, four product categories are listed, all sharing the same dataset:
- Caller ID API: a single GET request returns a spam determination, along with complaint evidence and business information behind the number, covering 195 countries. The returned content includes spam status scores, reputation and complaint history; business identity information such as name, location, and industry; carrier data such as carrier, number type, and number portability information, with optional HLR retrieval via the carrier lookup API.
- Spam phone number database Feed: covers reported numbers in 195 countries, retains the most recent 15 days of data, and is hosted on the customer side. It charges a fixed fee, has no per-query billing, zero latency, and can be refreshed at minute-level intervals. It is suitable for blocking at the switch layer, use as a dynamic do-not-originate list, and layering externally sold protection services on top of it.
- Call blocking SDK: about ten lines of code are enough to integrate spam protection into your own app. MVNOs and security vendors can sell it as a paid plan tier to increase ARPU. Materials state that queries on iOS and Android are anonymous: iOS uses Apple Live Caller ID Lookup, employing private information retrieval and homomorphic encryption; Android uses k-anonymity, so the HexPay side does not receive the queried number, avoiding leakage of subscriber call records. The SDK is fully white-label and supports Swift, Java/Kotlin, Flutter, and React Native, and also includes an Asterisk module.
- DID reputation monitoring API: a single call compares a number against 12 reputation sources and returns which of them flagged the number, making it easier to determine whom to contact for appeal and remediation. Wholesale carriers, CPaaS and UCaaS can embed it as a paid feature in their own dashboards. The material states that currently no other vendor provides this capability natively via API. In addition, the material mentions that fraudulent traffic can be dropped at the network edge, preventing it from reaching switches or consuming minutes, and can be used in conjunction with the Feed and API.
[Usage and Integration] The Caller ID API is self-service; after registering and obtaining an API key, the first query can be completed within minutes, and a free incoming call identification test is provided. Official ready-to-use integrations are provided for GoHighLevel, Aircall, and Dialpad, but such integrations are licensed through direct commercial agreements rather than self-service registration, and the team must be contacted for pricing based on deployment scope. Volume plans, Feed, and SDK transactions are all directly negotiated. SDK licensing is either a fixed fee or billed monthly per subscriber with call blocking enabled.
[Applicable Scenarios]
- CPaaS/UCaaS: improve ARPU with a call protection tier and reduce users answering scam calls;
- Voice AI platforms: screen before answering, reducing model and telephony costs for invalid calls;
- MVNOs and mobile security vendors: launch call blocking capability under their own brand as a paid plan tier;
- Wholesale carriers: monitor DID reputation for their own number inventory as a paid dashboard feature;
- Scenarios requiring zero-latency blocking at the switch or carrier level can use the self-hosted Feed.
[Coverage and Data Timeliness] Business phone number identification covers 195 countries; fraud and nuisance call detection currently covers 30 countries, and coverage continues to expand. The Feed updates every minute and includes a 15-day rolling nuisance history; numbers that are no longer active are phased out over time, while active nuisance numbers remain blocked.
[Pricing Information (as listed in the material)] The self-service Caller ID API plans start at $25, advertised as scalable to billions of calls. The material lists the Basic plan as including 5,000 queries per month, nuisance detection and business caller identification, and no-code integration; the Premium plan as 20,000 queries per month, including everything in Basic plus priority support; and there is also a pay-as-you-go option marked at $997, including 250,000+ queries, everything in Premium, and scalable pricing. When monthly queries exceed 300,000, volume plans start at $1,000 per month and can cover up to more than 50 million queries per month, priced by tier, with a quote provided after traffic details are supplied. The material does not give a uniform figure for specific Feed and SDK licensing prices; these must be negotiated directly.
[Customers and Confidentiality Practices] The material states that multiple deployments are under confidentiality agreements, and the only publicly disclosedFor anonymized cases: MVNOs, UCaaS, and platforms with marketplaces resell their nuisance protection and call intelligence under their own brand and profit from upgraded subscribers; a voice AI CPaaS platform encountered a coordinated fraudulent call campaign, and screening intercepted it before calls reached agents or consumed minutes; some niche MVNOs provide robocall blocking to subscribers under their own name, driven by a minute-level new Feed, with HexPay remaining invisible.
[Limitations and caveats] The materials do not provide specific terms such as free quotas, refunds, SLAs, or data processing agreements; nuisance detection country coverage (30) is significantly smaller than enterprise number identification coverage (195); GoHighLevel, Aircall, and Dialpad integrations do not support self-service activation and require a commercial agreement; for SDK technical details and sample code on each platform, the materials only give platform names and a code line count overview, without complete code.
[Key FAQ points (according to the materials)] The Caller ID API is for single-call queries and includes an enterprise number database for 195 countries and nuisance status; the Spam Database Feed contains only nuisance data and is a customer-hosted blacklist, refreshed every minute and retaining a 15-day rolling history, suitable for zero-latency blocking at the switch or carrier level. DID reputation monitoring compares 12 reputation sources in one request and returns the source of the flags. Data resale is explicitly permitted, meaning most MVNO, CPaaS, and UCaaS customers sell the protection capabilities as value-added services under their own brand.