Skip to main content
Deprecations:
  • Support for the Completions API for custom LLMs has been deprecated. The full removal of this support is planned for December 2026. Refer to your LLM providerโ€™s reference to switch to the appropriate API before the removal date.
  • Support for Googleโ€™s gemini-2.0-flash and gemini-2.0-flash-lite models has been deprecated. The full removal of this support is planned for June 1, 2026. Switch to the newer models, such as gemini-2.5-flash and gemini-2.5-flash-lite, before the removal date.
Help us improve our product documentation on docs.cognigy.com by sharing your thoughts in a quick survey. Your feedback shapes the future of our content!

2026.6.4

Cognigy

Released April 17, 2026

Bug Fixes

  • Fixed the issue where users couldnโ€™t access the Simulator in a Project due to an internal infrastructure error

2026.6.3

Cognigy

Released April 7, 2026

Bug Fixes

  • Fixed the connection between Cognigy and Voice Gateway, ensuring Mute Speech Input prevents speech barge-in. This update complements the fix from Voice Gateway patch 2026.6.2. To resolve this issue, make sure you update Cognigy to this version and Voice Gateway to 2026.6.2

2026.6.2

Cognigy

Released March 25, 2026

Improvements

  • Added redaction limit. Log entries with data payloads over 2.5 MB are no longer redacted and now show a placeholder message
  • Improved internal log redaction for PCI compliance

Cognigy Voice Gateway

Released April 7, 2026

Bug Fixes

  • Fixed the issue where activating the Mute Speech Input setting didnโ€™t prevent speech barge-in

2026.6.1

Cognigy

Released March 18, 2026

Bug Fixes

  • Fixed the service configuration to ensure correct deployment and operation

Cognigy Voice Gateway

Released March 31, 2026

Bug Fixes

  • Fixed the issue on the Login page in the Voice Gateway Self-Service Portal. The page displayed v2026.5.x as the current version even after the v2026.6.0 release. Now, the page displays the correct version

2026.6.0

Cognigy

Released March 17, 2026

Features

New Multimodal Agent Widget: Click To Call
Click To Call lets you easily embed a widget in your website or build a custom application that connects to a voice AI Agent for multimodal communication. With Click To Call, users can talk to your AI Agents directly from your website or application with a single click.
New Models from OpenAI and Microsoft Azure OpenAI
Introduced support for the gpt-5-1 model from OpenAI and Microsoft Azure OpenAI.
New Models from Anthropic
Introduced support for the claude-sonnet-4-6 model from Anthropic.

Improvements

  • Added a Language Configuration toggle to the AI Agent wizard to stop automatic language detection and prevent unintended switching
  • Added OAuth2 authentication support to the Cognigy API
  • Deactivated Knowledge Connector actions when the required Extension is missing
  • Added an error message when LLMs reach the token limit
  • Added a safety context preamble to Simulator prompts, reducing content policy errors when using Azure OpenAI with personas or missions that contain terms flagged by Azureโ€™s content management filters
  • Removed the beta tag from the Data Redaction settings section
  • Renamed the Sentences API tag to Example Sentences for clarity

Bug Fixes

  • Fixed the issue where the Management UI flickered when users were navigating between pages and scrolling content
  • Fixed the issue where the error message caused by LLMs reaching the token limit didnโ€™t provide the correct information
  • Fixed the issue where the Save button in the schedule configuration in the Simulator remained inactive without displaying any validation error message
  • Fixed the issue where the sourceName value for scheduled sources in the Input object included the _scheduled postfix
  • Fixed the issue where uploading Snapshots to Projects that didnโ€™t include a global LLM caused cache issues
  • Fixed the issue where setting the maximum tokens above 4000 in the LLM Prompt Node for Microsoft Azure OpenAIโ€™s gpt-4o and gpt-4o-mini models threw an error, and similar limits for Anthropicโ€™s claude-opus-4-0, claude-sonnet-4-0 and AWS Bedrockโ€™s claude-3-5-sonnet models
  • Fixed the issue where setting the Transcript Turns parameter to 0 in the LLM Prompt Node transcript didnโ€™t have any effect
  • Fixed the issue where Interaction Panel calls dropped if there was a Set Session Config Node in the Flow setting Microsoft provided by NiCE as TTS or STT vendor
  • Fixed the issue where the wrong schedule date was displayed for Knowledge syncs when scheduling was previously deactivated
  • Fixed the issue where the legacy privacy_policy option was displayed in the Profile Schema field in the Update Profile Node
  • Fixed the issue preventing the import of Packages with Knowledge Chunks created using custom models
  • Fixed the issue where the Last Edited columns on the Connector Configuration displayed edits that didnโ€™t take place
  • Fixed the issue where the Knowledge Connectors didnโ€™t work in trial environments
  • Fixed the issue where the error responses for the POST /persona/from-script/ API method used message instead of detail and didnโ€™t provide the correct data

Cognigy Voice Gateway

Released March 17, 2026

Bug Fixes

  • Fixed the issue where the Register status wasnโ€™t updated correctly on the Carrier page in the Voice Gateway Self-Service Portal due to a race condition in the outbound authentication process

Cognigy Live Agent

Released March 17, 2026

Bug Fixes

  • Fixed the issue where the avatar image wasnโ€™t aligned properly
  • Fixed the issue where notifications for conversations escalated to supervisors werenโ€™t translated

Cognigy Insights

Released March 17, 2026

Improvements

  • Redesigned the search logic in the Transcript Explorer to improve filtering and sorting against data in MongoDB
  • Removed the default order-by condition in the OData query
  • Added a lower timestamp bound to optimize deletion queries and avoid timeout errors

Bug Fixes

  • Fixed the issue where sorting and filtering in the Transcript Explorer didnโ€™t work correctly
  • Fixed the issue where the PATCH /v2.0/analytics API method failed while updating the PostgreSQL records due to a schema mismatch
  • Fixed the issue where the GET /v2.0/conversations API method didnโ€™t work correctly and returned missing, incorrect, or null rating values
  • Fixed the issue where messages sent through the WhatsApp Endpoint werenโ€™t displayed in the Transcript Explorer when the payload didnโ€™t contain top-level text

Infrastructure

Version Compatibility Matrix

The version compatibility matrix was updated for the following Cognigy products:
Last modified on November 3, 2026