Skip to main content
The DialNexa changelog for June 2026 summarizes the latest platform, API, voice AI, workflow, and documentation updates for teams building on DialNexa. Use this page to see what changed, what to review, and which documentation pages explain the current behavior.

Documentation Index

Fetch the complete documentation index at: https://dialnexa.com/docs/llms.txt AI agents and IDE assistants can connect to the docs MCP server at: https://dialnexa.com/docs/mcp Use this file to discover all available pages before exploring further.

June 2026 update summary

DialNexa changelog for June 2026

29th June, 2026

Review the sections below if you maintain Agent Builder prompts, inbound phone routes, outbound readiness, batch recipient uploads, transcriber selection, Speech to Speech agents, or post-call analysis reporting.

Agent Builder AI prompt editing

Added guidance for editing an existing draft prompt with DialNexa AI, including natural-language edit instructions, streamed prompt updates, word-level prompt diffs, chat history, and the review and test step before publishing. Related docs: Agent Builder, Testing Agents

Inbound call ringback during setup

Documented that Plivo inbound callers can hear ringback audio while DialNexa resolves the assigned agent version, prepares call context, starts recording, and redirects the call into the voice stream. Related docs: Inbound Calls, Phone Numbers

KYC verification and outbound readiness

Documented the current KYC workflow, including owner-facing banners, company legal name entry, registration and tax proof uploads, pending review, existing compliance reuse, rejected submission resubmission, and pre-verification outbound limits. Related docs: Phone Numbers, Batch Calls overview

Duplicate recipient handling

Added the default duplicate phone number skip behavior for batch recipient uploads, including first-row retention, skipped row messaging, empty phone row validation, and when to turn the setting off. Related docs: Common upload errors, Destination validation errors

Soniox STT RT v5 and language hints

Updated Soniox transcriber references to Soniox STT RT v5 and clarified that Hindi-English agents send valid en and hi language hints to Soniox. Related docs: Speech to text, Multilingual calls

Speech To Speech language choices

Added the Speech to Speech language catalog guidance for OpenAI realtime and Gemini model paths where enabled, including Indian languages, European languages, Hindi-English, and multilingual choices. Related docs: Agent stack controls, Provider selection guide

Post-call DateTime result display

Documented how DateTime post-call analysis fields remain ISO 8601 values with timezone information for exports while the call detail page can render them as readable date and time values. Related docs: Call detail page, Post-call analysis setup

v1 API endpoint coverage

Added API reference coverage for the v1 outbound-calls path, plural webcalls paths and create-response fields, nested agent-function routes, webhook-secret management, phone number orders, default number assignment, SIP trunk linking, Speech to Speech voices, and per-voice language choices. Related docs: Create a Web Call, Create Agent Function, Rotate Webhook Secret, List Speech To Speech Voices

22nd June, 2026

Review the sections below if you maintain knowledge bases, agent runtime settings, workflow automations, batch recipient uploads, or voice provider behavior.

Knowledge base management and retrieval settings

Updated knowledge base docs for workspace collections, document actions, agent version attachments, Chunks Retrieved, Similarity Threshold, and supported cascaded-agent retrieval behavior. Related docs: Managing documents, Using knowledge in agents

Agent time awareness and DateTime extraction

Added the agent timezone control and DateTime post-call analysis guidance so callback, appointment, renewal, and follow-up times can be interpreted consistently. Related docs: Post-call analysis, Workflow Time nodes

Workflow Application and Time node runtime behavior

Documented cloud Application node fields, connected integration actions, success and failure outputs, Composio execution billing, and PCA-based Time node scheduling. Related docs: Using integrations in workflows, Billing for cloud actions

Batch recipient CSV mapping

Updated batch upload docs for the Match CSV Columns modal, phone number auto-detection, dynamic variable matching, default values, unmapped field warnings, and grouped validation errors. Related docs: Common upload errors, Destination validation errors

Voice runtime and call validation updates

Refreshed voice and call docs for Soniox, fallback STT behavior, TTS ordering, web call warmup handling, phone destination validation, no-answer outcomes, and dynamic custom-function auth headers. Related docs: Web calls, Phone numbers, Custom functions

8th June, 2026

Review the sections below if you maintain guided agent creation flows, Speech to Speech agents, billing operations, workspace webhooks, workflow activation, or call monitoring workflows.

Agent Builder guided create flow

Added documentation for the guided Agent Builder flow, including structured intake, generated objectives and tone, streamed build phases, clarification questions, quality review, dynamic variable defaults, and draft agent creation. Related docs: Types of agents, Agents overview, Dynamic variables

Gemini Speech to Speech model support

Documented Gemini 3.1 Flash Live as a Speech to Speech model path, including Gemini voices, direct speech pipeline behavior, tool support, automatic activity detection, interruption handling, long-call continuity, welcome startup behavior, and current pricing preview expectations. Related docs: Speech to Speech agents, Speech to Speech models, Provider selection guide

Billing wallet and rate previews

Updated billing guidance for organization-scoped INR wallet balance, wallet transactions, rate previews, legacy billing behavior, telephony network group rates, SIP/BYOC flat rates, and super-admin credit adjustments. Related docs: Billing wallet behavior, Workspace settings tabs

Workspace webhook settings

Added the current Developer > Webhooks settings flow, including webhook URL, active status, friendly name, signing secret controls, retry count, timeout range, and the Update webhook action. Related docs: Workspace notification webhooks, Workspace settings tabs

Workflow activation validation

Expanded workflow docs for pre-activation validation, blocking graph and dependency errors, agent-change confirmation, workflow variable fallback checks, and active workflow conflict handling. Related docs: Workflow variables and activation, Workflow control actions

Call History filter and detail controls

Updated Call History docs for multi-agent, multi-status, and multi-sentiment filtering, page-size control, duration filtering in minutes, relative dates, keyboard detail navigation, and the resizable detail drawer. Related docs: Call status filters, Call detail page

1st June, 2026

Concrete status reference

Expanded status documentation for call logs, batch campaigns, workflow containers, workflow leads, Voice Call node routing, webhook runtime events, warm transfers, phone numbers, and integrations. Related docs: Batch call statuses, Workflow statuses, Call History status filters

API versioning and legacy route migration

Documented https://api.dialnexa.com/v1 as the recommended base URL for new integrations while keeping the legacy unversioned API documented for existing integrations until July 31, 2026. Related docs: Authentication, API errors, Pagination

Speech to Speech agent stack

Added guidance for Speech to Speech agents, including the create-agent choice, realtime model selector behavior, hidden transcriber controls, hidden voice model controls, Audio Cache exclusions, and lower max call duration. Related docs: Speech to Speech agents, Create agent API, Update agent API, Voice AI overview

Fallback STT settings

Documented fallback STT for cascaded agents, including the settings popover, fallback transcriber selection, fallback wait timing, and current Deepgram and Soniox selector behavior. Related docs: Supported transcribers, Agent stack controls

Audio Cache defaults

Updated Audio Cache guidance to explain that cascaded agents use Audio Cache by default, non-super-admin users should contact DialNexa support to disable it, and Speech to Speech agents do not use the separate TTS cache. Related docs: Provider selection guide, Glossary

Phone number nicknames

Added phone number nickname guidance for dashboard selectors, phone number lists, batch call setup, workflow call nodes, and API update requests. Related docs: List phone numbers, Get phone number, Purchase a phone number

Call end reasons

Expanded canonical end_reason documentation so call logs, webhooks, troubleshooting, and the call data model use the same values for user disconnects, busy lines, unreachable users, destination blocks, provider failures, capacity limits, and insufficient balance. Related docs: Call logs API, Call data model, Webhooks, Debug call issues

Batch recipient error CSV

Documented the Download errors CSV flow for recipient uploads with more than three validation failures, including the report columns and single-use download behavior. Related docs: Common upload errors

Workflow Time node parsing

Updated Time node docs for post-call analysis scheduling, including timezone resolution and accepted callback formats such as relative dates, named weekdays, explicit dates, comma-style month names, and slash dates. Related docs: Workflow builder concepts, Post-call analysis results

Workspace invite acceptance

Added the invite acceptance flow for /invite, including signed-out users, invited-email matches, wrong-account sign-ins, existing members, and expired or invalid links.