Three changes stand out this week, and two of them are Genesys removing things: a billing meter and a tuning knob. Both removals are good news. Here's my read on the July 13, 2026 release notes.
TL;DR
- Journey Management no longer charges for event overages. Ingestion is part of the base offering now; the per-agent event allotments and tiered overage fees are gone.
- The transcription confidence threshold setting is going away. The default drops to 0 and the control disappears from Speech and Text Analytics settings, so every transcription segment is kept.
- The Genesys TTS Connector is generally available on AppFoundry. You can plug third-party text-to-speech providers like Cartesia and ElevenLabs into Architect flows under the BYOT model.
Data, Analytics & Reporting
Journey Management: event overage charges removed
Until now, Journey Management billing came with an event budget tied to your seat counts (on the order of tens of thousands of events per agent per month), and blowing past those fair-usage thresholds triggered tiered overage charges. So someone on your team had to watch event volume and explain the surprise line item after a busy month.
That whole mechanism is gone. Event ingestion is now just included in the base Journey Management offering. No event-based fair usage limits, no overage billing. The one guardrail left is on the analysis side: query usage is still governed by the existing limits (Journey Analyzer has historically been capped at a set number of query analyses per day per org). This changes ingestion billing only; the analysis limits didn't move.
Transcription confidence threshold configuration removed
Speech and Text Analytics used to let you set a confidence threshold below which transcription segments got filtered out. Genesys is retiring that control: the default drops to 0 and the setting comes off the Speech and Text Analytics Settings page entirely.
The reasoning holds up. Transcription engines are accurate enough now that confidence filtering discards more signal than noise, so from here on every segment is preserved regardless of score. You get fuller transcripts. You also lose a subtle failure mode where sentiment analysis and topic spotting silently missed content because someone set a threshold years ago and forgot about it.
Open Platform
Genesys TTS Connector general availability
The Genesys TTS Connector is now generally available and installable straight from AppFoundry. It's a vendor-agnostic bridge: if you already pay for a supported third-party text-to-speech provider (Cartesia and ElevenLabs are the named examples), you can wire that voice into your Architect flows instead of settling for the natively integrated engines.
It follows the Bring Your Own Technology model and bills at the BYOT-A rate. You pay Genesys a platform rate; your provider contract and its pricing stay your own. Two caveats from the product docs: you have to configure both the Genesys side and the provider side, and the connector is not PCI-compliant. No connector-driven TTS in secure flows, full stop.
What this means for your contact center
The Journey Management change is the sleeper here. Event-based pricing was an adoption blocker, plain and simple: teams instrumented fewer touchpoints than they wanted because every event carried a marginal cost and somebody had to babysit the meter. With ingestion included, go back and look at what you didn't instrument, whether that's web events you skipped or custom milestones you trimmed to stay under the allotment, because the reason you held back no longer exists. Query limits still apply, though. Heavier instrumentation doesn't buy you unlimited analysis.
The transcription change is low-drama for most orgs. It matters if you ever customized the threshold. If you deliberately set a high confidence bar to keep noisy transcripts out of downstream tooling, that filter is gone, and any process that consumed transcripts assuming low-confidence segments were pre-filtered will now see them. Expect transcripts to run a little longer and occasionally messier.
The TTS Connector GA is the most strategically interesting of the three. The synthetic voice market moves much faster than any platform's native offerings, and a vendor-agnostic connector lets you swap providers as quality improves without rebuilding flows. I like it. But model the BYOT-A rate plus your provider's per-character pricing against native Genesys TTS before you commit, and remember the PCI limitation: secure flows stay on native engines.
What to check
- Confirm the Journey Management billing change is reflected on your account, retire any event-volume monitoring or alerting you built to dodge overages, and revisit instrumentation plans you scoped down.
- Check whether your org had a non-default transcription confidence threshold, and warn downstream consumers of transcripts (QA tooling, BI pipelines) that low-confidence segments will start appearing.
- Evaluate the Genesys TTS Connector on AppFoundry if you've wanted a specific third-party voice, and verify none of the target flows are secure flows, since the connector isn't PCI-compliant.
- Compare BYOT-A billing plus your third-party TTS provider's costs against your current native TTS spend before rolling the connector out broadly.
Full details are in the official Genesys Cloud release notes for July 13, 2026. Questions about how any of this lands in your environment? Email me.