Quick answer: Twilio maintains a public deprecation log, and the entries that matter to a CPaaS buyer are no longer limited to endpoint-level maintenance. Twilio Video — its WebRTC video API — is announced to sunset by the end of 2026, and Programmable Fax has been moved into the deprecation flow. Meanwhile the core Voice, SMS, and email (SendGrid) surfaces keep their own scheduled endpoint deprecations. The right response is not panic-rebuilding: it is budgeting a migration runbook per product family, with a product-mapping page per family, so a sunset becomes a scheduled swap rather than an emergency. On Devotel Orbit, the runbook maps cleanly to the shipped voice, video, and SMS surfaces, with the public migration docs as the step-by-step spine.
If you are a current Twilio customer, the question is not "is my product going away" — it is "which product family owns which line of my table, and what does the per-family migration path look like?" That is the budgeting frame this explainer applies.
Which announced Twilio deprecations actually affect your runbook
Vendor deprecations land in three different severities, and treating them as one class is what makes a migration budget wrong. Sort the log first:
- Product-line sunsets. The entire product is retiring. Twilio Video is the current flagship entry — a shipped product line with a full-sunset date by end of 2026, not a version bump. On the messaging side, Programmable Fax has been moved into the deprecation flow. A product-line sunset is a "move the family" event: pick a replacement surface, port the usage, and retire the Twilio dependency.
- Endpoint-level scheduled deprecations. Individual API versions, REST resources, or webhook shapes retire on a published date. The Twilio Voice and Programmable SMS APIs, and the SendGrid email surface, publish these regularly. These are maintenance events — schedule the SDK bump, update the payload mapping, rerun the integration tests, and move on.
- Behavioral changes without a deprecation label. A pricing change, a deliverability policy update, or a rate-limit adjustment can change behavior with no sunset notice at all. These do not appear in the deprecation log, which is why a "vendor-risk review" should include the pricing page and the trust/policy center, not only the deprecation feed.
The budgeting consequence: product-line sunsets get a work-stream (a project with a pick-a-replacement decision), endpoint deprecations get a maintenance line item (a scheduled engineering task), and behavioral changes get a monitoring line item (a review cadence). A sunset that arrives as "we should have revisited Voice three years ago" costs far more than the same event planned into a quarterly review.
What a migration runbook looks like per product family
A runbook is a checklist a team can execute on a clock. For a vendor deprecation, one runbook per product family keeps the scope honest — a sunset does not get smuggled across all families at once.
Voice. Map the current telephony surface: which numbers are live, which flows (IVR, queue, recording, AI agent) consume them, and which outbound campaigns terminate on the route. The migration runbook ports numbers, rebuilds the flow logic, and re-verifies call quality against the new termination path. On Orbit the programmable voice API is the corresponding family, and the phone number porting guide covers the numbering step.
SMS. Map the messaging surfaces: the sender IDs, the webhook endpoints, and the per-campaign routing. The runbook ports numbers, moves webhook consumers, and re-registers any regulated campaign identifiers (10DLC in the US). On Orbit the SMS API is the corresponding family; sender-level concerns are tenant-owned configuration, not a platform-level accidental change.
Video. For a product-line sunset like Twilio Video, map which realtime sessions run on the retiring product — embedded video support, in-app calls, or recording pipelines. The runbook picks a replacement realtime surface and re-points the session logic. On Devotel Orbit the video API is that surface, covering scheduled rooms, server-minted join tokens, recording, and moderation, on the same account as voice and SMS — so the choice is a feature checklist, not a rebuild.
Fax and email. Fax has a dedicated surface on Orbit (the fax API for healthcare and legal workflows covers the fax family). Email maps to the HTTP API surface a team already operates — on Orbit the email channel is a first-class messaging family. In both cases the runbook ports sender identities and webhook shapes, not whole product paradigms.
Each family runbook ends with a verification pass: a smoke test against the new endpoint, a rollback trigger, and a monitoring subscription (Orbit's anomaly-reporting surfaces are tenant-visible signals, not vendor-monitoring black boxes). The sequence order matters — port before you retire, verify before you cut over, and keep the old vendor in parallel only as long as the plan explicitly allows.
How the runbook pattern maps to /voice-api, /video-api, and the public migration docs
The point of separating product families is that each one maps to a concrete destination. On Orbit, voice and SMS map to their own platform surfaces, and a realtime-video sunset maps to the video API. The public migration from Twilio guide is the step-by-step spine for the voice and SMS families — endpoint mapping, number porting, webhook migration — written in the same sequence a runbook follows.
The alternative to reading this as a deprecation-only topic: when your platform is consolidated, a deprecation is a planned event with a known family boundary, not an architecture review. That is the same logic the best Twilio alternatives for voice, SMS, and AI agents ranked guide applies — it scores providers on shipped, checkable capability — and the Orbit vs Twilio comparison puts the per-family mapping in one place, so the "pick a replacement" decision a sunset triggers is a page read, not a quarter-long evaluation.
Frequently asked questions
Which Twilio product deprecations should a CPaaS buyer watch first?
The product-line sunsets — Twilio Video's announced full sunset by end of 2026, and Programmable Fax entering the deprecation flow — are work-stream events. The scheduled endpoint deprecations on Voice, SMS, and SendGrid email are maintenance events. Both classes belong in a per-quarter vendor-risk review.
What does a migration runbook per product family look like?
One runbook per family (voice, SMS, video, fax, email). Each maps the current surface — numbers, flows, senders, sessions — then ports, verifies, cuts over, and monitors. The video sunset maps into a WebRTC replacement; the SMS and voice deprecations map into the programmable voice and SMS surfaces on the platform you pick.
How does the runbook pattern map to Orbit's voice and video APIs?
Voice and SMS map to the programmable voice and SMS API surfaces directly. A WebRTC video sunset maps to the video API — scheduled rooms, join tokens, recording, and moderation on the same account as voice and SMS. The migration from Twilio guide covers the porting sequence step by step.
Where do I find the ranked shortlist of Twilio alternatives?
The best Twilio alternatives for voice, SMS, and AI agents guide scores providers on shipped capability. The Orbit vs Twilio comparison page shows the per-product-family mapping in one place, so the evaluation step a sunset forces stays short.
The takeaway
Twilio's deprecation log is no longer only endpoint maintenance — it now includes a product-line sunset in Twilio Video and a Programmable Fax deprecation. Split the log into product-line sunsets, endpoint deprecations, and behavioral changes; assign each a work-stream, maintenance, or monitoring line item. Then read the runbook one product family at a time, mapping each to a concrete replacement surface. If your platform choice is still open, the best Twilio alternatives guide is the ranked starting point, and the migration from Twilio docs are the step-by-step path on Orbit.