Ink

Last update: September 9, 2026

Conversations and Inbox

Call monitoring for supervisors [#call-monitoring-conversations-and-inbox]

Supervisors can now join an in-progress call as a silent listener, without the agent or customer being aware. This gives supervisors real-time insight into how agents handle calls, supporting quality assurance and coaching.

Key details:

• Available to users with the Supervisor role.
• Access it from Insights and Analytics > Analytics > Live analytics, by selecting Listen next to a call in the Active calls table or next to an agent in the Human agents table (only available while that agent is on a call).
• Only one supervisor can monitor a call at a time.
• Monitoring continues through warm transfers, but is not available for calls with external participants.

For more information, see Call monitoring in Inbox and Call monitoring in Conversations.

New "Assignment changed" workflow trigger [#assignment-changed-conversations-and-inbox]

Workflows now support an Assignment changed trigger, which fires whenever a conversation is assigned, unassigned, or reassigned to a different agent.

Combine it with the Agent condition to react to a specific type of assignment change. For example, set the condition to Assigned to send a personalized welcome message once a conversation is assigned to an agent, or to Unassigned to react only when a conversation becomes unassigned.

For more information, see Triggers in Inbox and Triggers in Conversations.

New capacity models for automatic assignment [#capacity-model-conversations-and-inbox]

You can now choose how agent workload is calculated for automatic assignment, using a new capacity model setting:

• Combined capacity: All channels share one workload threshold. When an agent's total load reaches the limit, they stop receiving new conversations on every channel, even the ones that are not full.

• Channel type capacity: Each channel type has its own limit. When an agent reaches capacity on one channel type, they can still receive new conversations on the others, including text conversations while on a call.

Key details:

• The capacity model is set at the account level and applies to all agents.
• The Full capacity status threshold is now 100% of an agent's maximum capacity (previously 90%), and how it is calculated depends on the selected model.
• Manually assigning conversations beyond an agent's capacity is not blocked and does not push channel occupancy above 100% in the calculation.
• Switching models reassigns waiting conversations under the new model; conversations already assigned are not affected.

For more information, see Capacity in Inbox and Capacity in Conversations.

Unified template creation flow for all channels [#unified-template-creation-flow-conversations]

Creating and managing templates is now consistent across every channel, not just WhatsApp.

What is new:

• The template status toggle is now available for templates on all channels. Previously, it was visible but disabled for non-WhatsApp templates, which were always active by default.
• Template creation now follows the same two-step flow for every channel: a content step, followed by a Setup step where you configure status, template validity, access and visibility, language, and tags.
• When editing an existing template, you can now choose to Save the changes or Duplicate the template into a new one.
• You can filter templates by validity: Any time, Currently available, or a custom date range.
• WhatsApp Click-to-URL buttons now support dynamic URLs, in addition to static ones. Add a placeholder for the customer-specific part of the link, and configure it to be editable by agents or prepopulated from a Customer Profiles attribute.

Key benefits

• Consistent, predictable template setup regardless of channel.
• Supervisors can control template availability and access on any channel, not only WhatsApp.
• More flexible WhatsApp templates with dynamic links for personalized customer journeys.

Improved template picker and preview for agents [#improved-template-picker-conversations]

Agents can now preview and use templates more efficiently, with the same improved experience across channels.

What is new:

• The template picker supports searching by name, content, or tag, and narrowing results by Tag and Language.
• Selecting a template shows a preview of its fully rendered content, including attachments, before adding it to a reply.
• Once selected, the template content and attachments load directly into the message composer, ready to review, edit, and send.
• Supervisors can review templates with missing content directly from the Templates list and choose to pre-fill placeholders or leave them editable by agents.

Key benefits

• Agents can confirm a template looks right before sending it to a customer.
• Fewer errors from missing media or unfilled placeholders.
• Faster, more consistent responses across all supported channels.

For more information on both Template updates, see Templates in Inbox and Templates in Conversations.

RCS

Launch to additional providers [#launch-to-additional-providers-rcs]

You can now request launch to additional RCS providers for senders that are already launched in a country. When a new provider becomes available, the sender status changes to Launched - Partial success, and you can submit a launch request for the additional provider through the web interface or API.

Previously, if a provider became available after a sender was launched, you had to contact support to request launch to the additional carrier.

Key details:

• Sender status changes from Launched to Launched - Partial success when an additional provider becomes available.
• Submit a launch request for the additional provider from the Senders page or through the API.
• The PROVIDER_UPDATE webhook event notifies you when an additional provider becomes available in a country.
• The SENDER_UPDATE webhook includes a mandatory reason field that explains the status change.

For more information, see Launch to additional providers and RCS over API - Sender update webhook.

Sender status transition improvements [#sender-status-transition-improvements-rcs]

The sender launch process is now more transparent. The sender is locked immediately when a launch request is submitted, and the status is set to Pending launch.

Previously, the sender was not locked immediately when a launch request was submitted, which could cause confusion about the sender state during the review process.

Key details:

• Sender is locked and set to Pending launch as soon as a launch request is submitted.
• If the launch requires sender edits, the status changes to Update required and the sender is unlocked.
• Status changes are visible in the web interface and API.

For more information, see Create your RCS sender - Sender statuses.

Expanded CONVERSATION_STARTED webhook [#expanded-conversation-started-webhook-rcs]

The CONVERSATION_STARTED webhook now includes sender, recipient, and direction fields, so you can identify conversation participants without making separate API calls.

Previously, the webhook did not include participant or direction information.

Key details:

• sender: the sender name (outbound) or end user phone number (inbound).
• to: the destination phone number (outbound) or sender name (inbound).
• direction: direction of the message: OUTBOUND (MT) or INBOUND (MO).

For more information, see RCS over API - Conversation Started webhook event. For API details, see the Receive RCS conversation started events API reference.

Reporting

Kakao costs in Financial reports and the Billing Usage API [#kakao-financial-reports]

Kakao channel costs are now available in Financial reports and the Billing Usage API. View Kakao spend alongside other channels, compare costs across your messaging activity, and reconcile Kakao charges with your billing data.

Key benefits:

• Compare channel spend: View Kakao costs alongside costs for other channels in Financial reports.
• Automate reconciliation: Query Kakao cost data with the Billing Usage API and integrate it into your reporting or reconciliation workflows.
• Analyze costs consistently: Use both reporting surfaces to review Kakao spend and distinguish it from costs for other messaging channels.

For more information, see Financial reports and the Billing Usage API.

Interaction type dimension for Kakao and RCS [#interaction-type-reporting]

The delivery and traffic performance report now includes an Interaction type dimension for Kakao and RCS. Interaction type describes the message format used, so you can group traffic and delivery data by message format for these channels.

Key details:

• Where it applies: Available for Kakao and RCS in the delivery and traffic performance report. See Report structure and grouping logic.
• Performance, not billing: Interaction type is a reporting dimension, not a billing category. Use the delivery and traffic performance report to compare message activity, and use Financial reports and the Billing Usage API to analyze costs. For documented billing classifications, see Traffic types by channel.
• RCS webhooks keep the concepts separate: trafficType identifies the billing classification, while interactionType identifies the interaction classification. Conversation fields provide session details. See Billing transparency in webhooks.

For more information, see the delivery and traffic performance report.

Clicks by URL report [#clicks-by-url-report]

The new Clicks by URL report shows engagement for each individual URL in messages that use URL tracking and shortening. Instead of seeing only the total number of clicks for a message, you can identify which links drove engagement and compare link performance across campaigns.

Key benefits:

• Measure link-level engagement: See which URLs receive clicks within a message.
• Compare calls to action: Use click data to evaluate different offers, destinations, and calls to action.
• Optimize campaigns: Identify the links that perform best and use those insights to improve future campaigns.

The Reports area is now unified. Reports and New reports are available in one Reports section with grouped secondary navigation, so you can access all reports from a single location.

For more information, see Clicks by URL report.

Resources and numbers

Structured launch feedback [#structured-launch-feedback-resources-and-numbers]

Resource Request now supports structured feedback. When enabled, you receive clearer guidance on what needs to be fixed and why, making it easier to update your request and move forward faster. Previously, feedback was shared as free-text messages. Now it is returned in a structured format that is easier to understand, act on, and use in automated workflows.

Key details:

• Set structuredFeedback to true in the Resource Request API launch request to opt in to structured feedback.
• Rejection feedback includes requirementKey (sender data fields) and externalRequirementKey (launch data fields).
• Each feedback entry contains an errors array with key, reason, and details.

For more information, see the Resource Request status change notification API reference and Launch your RCS sender.

Viber

GIF support for image messages [#gif-support-image-messages-viber]

Viber Business Messages image messages now support animated and non-animated GIF files. You can include GIFs in image message templates alongside the existing supported formats: .bmp, .jpg, .jpeg, .png, .svg, and .webp.

Previously, GIF files were not listed as a supported format for Viber Business Messages image messages.

Key details:

• Supported for Viber Business Messages image messages.
• GIF files allow up to 24 MB.
• Use HTTPS URLs when referencing GIF files in image messages.

For more information, see Viber Business Messages - Image messages.

Voice AI agents

Customer silence and re-engagement handling [#ink-voice-ai-agents-customer-silence-re-engagement]

Voice AI agents now re-engage proactively when a caller goes quiet. When a caller stops responding, the Voice AI agent sends a re-engagement prompt to keep the conversation moving. If the caller stays silent after a set number of retries, the Voice AI agent ends the call or returns control to the journey, so every call reaches a clear outcome.

What is new:

• Timeout: Set how long to wait for the caller before the Voice AI agent re-engages.
• Retries after timeout: Set how many times the Voice AI agent re-engages before ending the call or returning to the journey.

Configure these in the Inactivity handling settings of the Transfer call to AI agent element. Inactivity handling is supported for both Speech to speech and Speech to text to speech agents.

For more information, see Speech to speech or Speech to text to speech section.

Contextual progress messages [#ink-voice-ai-agents-contextual-progress-messages]

Voice AI agents can now speak contextual progress messages while processing a request, so callers always hear what the Voice AI agent is doing. For example, the Voice AI agent can say "Let me check the available times" while it looks up a booking. Static messages remain available as the default.

What is new:

• Progress message strategy: Choose how the Voice AI agent handles processing pauses.
• Prompt instructions: Shape the contextual messages the Voice AI agent generates.
• Music file: Play a background sound while the Voice AI agent processes the request.

Available for Speech to text to speech agents in the Transfer call to AI agent element.

For more information, see Speech to text to speech section.

Voice and Video

Number reputation scoring [#voice-number-reputation]

Number reputation scoring is now available on the Infobip web interface, powered by Hiya. Monitor the reputation of your voice calling numbers and identify numbers that carriers or end users have flagged as spam or fraud.

Use number reputation data to maintain healthy calling numbers, improve answer rates, and take corrective action before reputation issues affect your campaigns.

Key details:

• View number reputation scores on the Infobip web interface under Voice trust center.
• Powered by Hiya reputation data.
• Register business numbers for reputation monitoring.
• Retrieve reputation scores through the API.

For more information, see Number reputation and the Number reputation page on the web interface. For API details, see the Number reputation Hiya API reference.

WhatsApp

Support template validity period [#whatsapp-template-validity-period]

You can now define how long a WhatsApp template message stays valid before it expires. This gives you more control over delivery timing for time-sensitive messages, such as booking confirmations, delivery updates, and limited-time promotions, so messages are not delivered after they are no longer relevant.

The validity period was already available for authentication templates. It now also applies to utility and marketing templates.

You set the validity period when you create a WhatsApp template or edit a WhatsApp template. The supported range, and the default that applies when you do not set a value, depend on the template category:

Key details:

• A validity period always applies. When you do not set one, Infobip applies the category default, so messages are not delivered after they are no longer valid.
• A validity period set during template creation overrides the default TTL defined by Meta.
• Marketing validity periods apply only to messages sent through the Marketing Messages API (MM API). You can still define a validity period on any marketing template, but it is ignored when the template is sent over the Cloud API.
• Available now through the API, with web interface support coming soon.