Migrating off Amazon Connect Voice ID
Voice ID has reached end of support. Here is how IdentityCall replaces speaker verification and identification, with an API and call analysis Voice ID did not include.
Last updated: September 2026
Amazon Connect Voice ID is end-of-life
According to AWS, Amazon Connect Voice ID closed to new customers on May 20, 2025 and reached end of support on May 20, 2026. Existing deployments must move to another speaker-recognition provider. Verify the current status in AWS documentation before you plan a cutover.
When AWS retired Amazon Connect Voice ID, its guidance listed Marketplace partners such as Pindrop and a DIY SMS one-time-password option. The latter replaces the authentication flow rather than providing voice biometrics. Pindrop does not publish self-serve pricing, so teams that adopted Voice ID for its usage pricing and API should compare quotes on the same usage assumptions.
IdentityCall is built for that gap. It delivers speaker enrollment, verification, and identification through a REST API, at published SMB and mid-market pricing, and adds the call analysis Voice ID did not include: transcription, QA scoring, categorization, emotion analytics, and compliance disclosure detection. You replace one capability and gain five.
Amazon Connect Voice ID (historical) vs. IdentityCall
| Feature | IdentityCall | Amazon Connect Voice ID |
|---|---|---|
| Status | Generally available | End of support (May 20, 2026) |
| Speaker verification |  | |
| Speaker identification | | |
| Works outside Amazon Connect | API, uploads, own numbers | Amazon Connect only |
| Transcription & diarization | | No |
| Automated QA / call scoring | | No |
| Emotion analytics | | No |
| Compliance disclosure detection | | No |
| AI-cloned voice detection | | Limited |
| Self-serve API & published pricing | | Previously available; service retired |
| EU / GDPR data handling | | AWS regions |
Competitor facts (Amazon Connect Voice ID end-of-support timeline and AWS migration guidance) are drawn from AWS documentation as of September 2026. The competitor column describes the service as it was before retirement. AWS may change guidance; re-verify against the current AWS docs before making decisions.
Choose IdentityCall if
- You are migrating off Voice ID and want an API-first replacement without an enterprise contract.
- You want speaker recognition plus transcription, QA, emotion, and compliance in one platform.
- You record or upload calls from outside Amazon Connect.
- You need EU/GDPR data handling and transparent pricing.
Consider an AWS-native path if
- •You are deeply embedded in Amazon Connect and need an in-console replacement.
- •You are an enterprise comfortable with quote-based contracts (for example, Pindrop).
- •Speaker biometrics is the only capability you need and call analysis is irrelevant.
Frequently asked questions
When does Amazon Connect Voice ID shut down?
Per AWS, Voice ID closed to new customers on May 20, 2025 and reached end of support on May 20, 2026. After end of support, applications can no longer rely on the Voice ID APIs. Confirm the current status in AWS documentation.
What does AWS recommend migrating to?
AWS lists Marketplace partners such as Pindrop and a DIY SMS one-time-password option. The latter replaces the authentication flow rather than providing voice biometrics. Pindrop does not publish self-serve pricing. IdentityCall is an API-first alternative with published SMB and mid-market tiers.
Can I transfer my existing voiceprints?
Do not assume existing speaker profiles are transferable. AWS has confirmed Voice ID enrollments cannot transfer to Pindrop, and other vendor pairs need checking. Plan for re-enrollment in IdentityCall, and confirm whether retained recordings are permitted for the intended use and meet the new provider’s requirements before rebuilding profiles from them.
Does IdentityCall require Amazon Connect?
No. IdentityCall works through a REST API, direct audio uploads, and its own virtual numbers, so it is not tied to any single telephony platform.
What do I gain beyond speaker recognition?
IdentityCall analyzes every call: transcription with diarization, goal-based QA scoring with reasoning, auto-categorization, per-segment emotion analytics, and compliance disclosure detection, alongside the biometrics you are replacing.
Plan your migration off Voice ID
Talk to us about moving speaker verification and identification to IdentityCall, with call analysis included.
