Our guides to texting agents and business SMS platforms are built from what vendors publish and what independent reporters and regulators have documented. We do not run hands-on tests, send production traffic through these platforms, or publish usage-based scores. Instead, we rate products on published facts and date every price and availability claim.

What sources do we use?

In order of preference:

  1. Official documentation, help centres and changelogs, for example Twilio's pricing and changelog, Meta's WhatsApp Business Platform pricing and vendor help articles.
  2. Official pricing pages for plan names, per-message and per-resolution prices.
  3. Regulator and legal sources: FCC and FTC material, the UK ICO and Ofcom, and law-firm summaries where primary orders were not available (we say when that is the case).
  4. Reputable press and trade coverage for launch dates, incidents and context.

Third-party pricing estimates, such as reported prices for tools that quote privately, are labelled as estimates. Social posts are used only as leads that we then try to confirm.

  1. CollectGather docs, pricing pages, changelogs and regulator material.
  2. Cross-checkCompare with press and trade coverage.
  3. Flag conflictsWhere sources disagree, show both and say which is more likely.
  4. RateAssign Strong, Mixed or Weak on each criterion.
  5. Date and reviseMark claims "as of" the month checked and update when things change.

What criteria do we rate?

Criterion Personal texting agents Business SMS platforms
Channel support SMS, iMessage, RCS, WhatsApp, Telegram, availability by country Channels the agent runs on, countries, number types
Price Free tier, plan prices, message limits Per-segment, per-resolution or per-credit pricing, fees
Capability Actions, memory, reminders, voice notes Agent tooling, memory, handoff, integrations
Safety Approval gates, prompt-injection defences, incident record Opt-out handling, verification, guardrails
Privacy Encryption, training defaults, deletion Data handling, BAAs or DPAs, data residency

Ratings are plain labels (Strong, Mixed, Weak) based on documented facts, not a single combined score, because the right choice depends on what matters to you.

How do we handle missing or conflicting facts?

When a vendor has not published something, such as AI pricing or a country list, we say it is not published rather than guessing. When sources conflict, such as whether a channel is live or "coming soon", we report both. Cost examples that we calculate ourselves from list prices are marked as our arithmetic, not quoted figures.

How often do we update?

Pricing, channel and compliance pages are reviewed whenever a vendor or regulator announces a change, and every page shows its last updated date. Read more about this site.