Changelog

Voicelink joins the SIP providers you can connect directly. Import your trunk, set a dial prefix or strip digits to match your carrier's format, and authenticate by IP without waiting on support.

Add it under Phone Numbers > Import SIP trunk. Full setup steps are in the telephony docs.

You can now subscribe to any plan annually. The full year of minutes lands in your balance the moment you subscribe, instead of arriving month by month.

Switch between monthly and yearly on the Pricing page. Each plan card shows the yearly minute total, and the plan-change dialog prices every option by its interval so you see the yearly or monthly cost before you commit.

Invoices are easier to read. Every invoice now shows your credit balance before and after it, and line descriptions spell out the plan, billing period, and top-up amount instead of a bare total.

Invoices still being generated appear as a processing row, so you know one is on the way instead of wondering where it went. Find them all under Billing > Invoices.

The OmniDimension MCP server, which connects your agents and calls to Claude and other MCP clients, picked up several upgrades:

• Tools are now labelled by what they do, so clients can tell read-only actions apart from ones that change or delete data before running them.
• Voice and language details are available as reference resources, making it easier to pick the right voice for each language.
• Failures are reported clearly, so a failed call no longer looks like it succeeded.

The TypeScript SDK types have also been refreshed to match the latest API.

OmniRelay partners can now make more of the experience their own:

• Branded documentation. Turn on a docs section under your own domain and add your own custom pages for your customers.
• Email from your domain. Connect your own outgoing email so messages to your customers come from your address.

Both are set up from your OmniRelay reseller settings.

Creating a WhatsApp message template that Meta rejected used to show a generic failure. The exact reason from Meta now appears in the error, so you can fix the template and try again without guessing.

Opening Phone Numbers used to slow down as you added more numbers. It now loads each number's connected agent in a single pass, so the page opens quickly no matter how many numbers you have.

Numbers on the older on-premise WhatsApp setup can't send message templates. You can now move a number to the WhatsApp Cloud API directly from the dashboard, with a guided flow that walks you through each step.

Open WhatsApp > Numbers. Any number that needs to move shows a badge with a link to Meta's instructions and a button to start the switch.

Fixed an issue where the automatic retry step could restart a bulk campaign you had already stopped. Stopped campaigns now stay stopped.

Fixed an issue where a call that finished without a post-call analysis could send the same notification email twice. These emails now go out exactly once.

Greeting is now generated on a fast, dedicated model decoupled from the main conversation LLM. Generation starts before the caller connects. A persistent LLM socket is warmed the moment call config arrives, eliminating cold-socket setup cost. Greeting is streamed to TTS sentence-by-sentence so first audio plays sooner. Main LLM is warmed during greeting playback so the first real reply is fast too. A first-audio latency budget bounds worst-case greeting delay.

Collaborative language model with sticky explicit requests, a per-turn language directive to the LLM, and a conversation-language state machine with backchannel hysteresis so a stray "yes/हाँ" no longer flips the language. Full language packs for all 13 dashboard languages with variant normalization, plus end-to-end hardening for Bengali, Tamil, Marathi, Telugu, and Gujarati. Transliteration-aware detection handles Soniox romanized output via romanized function-word packs and phonetic cross-script matching. Negation-aware switch requests — "don't speak Hindi" is no longer treated as a request for Hindi.

Self-serve white-label setup wizard covering branding (light and dark logos) and custom domain with DNS guidance (CNAME/A, apex-aware). New agency pricing tier with a billing tab showing plan-usage card and client-org caps with upgrade CTA at the cap. Reseller custom SMTP lets Premium tier agencies send platform email from their own mail server. New /agency-owners landing page.

Revamped call-insight charts with a weekday heatmap. Connect-rate now uses Wilson lower bound with a best-call-window suggestion. New Total Calls Made and Connected Calls columns with deltas labeled vs previous N days. Dashboard loads are noticeably faster thanks to parallel section queries and skeletons.

USD-native top-ups are now available for users outside India.

Voicemail detection status is now shown in call logs with a dedicated voicemail filter. Hangup reason now includes system and carrier "ended by" labels so you know exactly why a call dropped.

MCP server updated from v0.5.1 to v0.6.0 with a new procedure layer (provision_agent and audit_calls guided prompts) and update-available notifications. MCP Cloud updated from v0.3.0 to v0.3.1 with procedure layer prompts and resources. TypeScript SDK v0.1.1 ships with types regenerated from the live spec.

A failed LLM attempt no longer triggers the pipeline kill path while failover is mid-recovery. Per-attempt failure signal and cooldown raised to 3 strikes for steadier recovery. Welcome-turn TTFT grace stops a pointless failover storm during the greeting.

Each node now tracks its own execution stats - run count, failure rate, and last execution time - visible directly in the workflow builder. For debugging, any node can now be tested in isolation using its live on-screen configuration without saving the workflow first. Test runs also collect a sample payload for every trigger type, so you always have real data to work with.

Four new nodes are live in the workflow builder:

Run Workflow - trigger a sub-workflow from within any workflow, with a depth cap to prevent infinite loops.

Notes - add inline documentation to any workflow; the node renders its text and passes it as output, so it doubles as a lightweight data node.

Slack - send messages to a Slack channel directly from a workflow step, no Zapier or Make required.

Code - run custom logic in a sandboxed eval environment. Use it for transformations, conditionals, or anything the standard nodes don't cover.

All four are additive — no existing workflows are affected.

The bot runtime has been restructured so the platform layer controls how the bot speaks - brevity, accuracy, persona - while the bot prompt controls what it knows and does. This eliminates conflicts between the two layers and makes multilingual switching reliable.

Key changes: the LANGUAGE block now defers to a per-turn conversation directive instead of self-detecting, which fixes mid-call language switching for Hindi, English, Telugu, and other configured languages. Names are title-cased automatically and the bot falls back to the first listed language when no directive is present.

New bot generation now produces lean, guide-aligned prompts under a 450-word cap across 7 structured sections - no TTS instructions, no language restatements, no closing scripts in the prompt itself. Languages are auto-selected at bot creation (default: Hindi + English) unless explicitly chosen.

When a node was duplicated or pasted, it inherited the same ID as the original. Any {{ context.<id> }} reference pointing to that node would silently break or resolve to the wrong value. Node IDs are now reminted on every duplicate and paste operation, so context references stay accurate across the workflow.

Install @omnidim-ai/sdk from npm and build on OmniDimension from your backend. Manage agents, dispatch calls, run bulk campaigns, knowledge bases, and phone numbers with typed methods and editor autocomplete.

It works in both TypeScript and JavaScript and runs server-side with your API key. Install with npm install @omnidim-ai/sdk.

See the TypeScript SDK guide to get started. The API reference now shows the SDK call for every endpoint.

Recording ingestion now runs as a dedicated background service. Recordings are stored immediately after a call ends, and slower carrier fetches are processed on a separate queue so they never delay anything else. If a recording arrives before the call record is ready, the system retries automatically.

All recordings are stored and accessible from the same place, regardless of carrier.

Post-call processing now runs as a dedicated background service. After a call ends, AI extracts a summary, sentiment, and any custom variables you've configured - then delivers the results to all connected destinations: Slack, email, webhook, CRM, WhatsApp, and Google Sheets.

Google Sheets delivery goes through a dedicated queue worker. Rows are written with full call metadata, new columns are added automatically as your data evolves, and token refreshes happen transparently. Any result that doesn't land is detected and retried automatically - no dropped data.

OmniDimension now runs as a Model Context Protocol (MCP) server, so you can manage your account from Claude and any other MCP client. List and edit agents, dispatch calls, run bulk campaigns, and read call logs without leaving the chat.

Connecting takes one sign-in. You approve access through a consent screen, and the client uses that grant to act on your behalf. Both a hosted cloud option and a local option are available.

See the setup guides under Docs > MCP at docs.omnidim.io.

We reworked how bulk campaigns run. Calls now dispatch in parallel, so large campaigns start faster, and the system no longer falsely marks active calls as failed under load.

Calls that get stuck mid-flight are now detected and recovered automatically, and you can retry failed lines with a single button.

Your billing now has its own home in the dashboard. The new tabbed page shows your current plan on the overview, a full invoice history, and the phone numbers on your account.

Every invoice can be downloaded as a PDF, and we email a copy when one is generated.

Open it from Billing in the dashboard.

The pricing page now has an Agency tab covering our whitelabel and reseller API options, so you can resell OmniDimension under your own brand.

Whitelabel partners can manage their own setup directly, including their privacy policy URL, billing link, and branding on customer-facing pages like email unsubscribe.

See the Agency tab on the pricing page to get started.

The integrations directory at omnidim.io/integrations got a full redesign. You can now search and filter across every supported integration, and each integration has its own detail page with setup steps, key benefits, common use cases, and video walkthroughs where available.

What's new:

• Detail page per integration. HubSpot, Salesforce, Twilio, Cal.com, Google Calendar, RingCentral, Vonage, Exotel, Zapier, Make, n8n, Slack, WhatsApp, Custom API, Webhook, and Web Search. Each page includes setup guides linked to docs.omnidim.io.
• Search and filter. Pill-style category filters plus a full-text search overlay for finding integrations by name or capability.
• Video walkthroughs. YouTube tutorial grids embedded on the detail pages (where applicable)
• Category structure. Calendar & CRM, Automation, Telephony, Messaging, Custom & Tools.

The complete OpenAPI 3.1 specification for the OmniDimension REST API is now downloadable as a single file at docs.omnidim.io/openapi.yaml.

What you can do with it:

• Import into Postman, Insomnia, or Stoplight for an interactive API workspace.
• Generate type-safe client SDKs in any language using openapi-generator or similar tools.
• Feed it to AI agents (Claude, ChatGPT, Perplexity) to let them call OmniDimension endpoints natively.
• Validate request and response shapes during development.

The spec stays in sync with the public REST API documented at docs.omnidim.io/docs/api-reference. All 49 endpoints are covered with full request, response, and schema definitions.

We rolled out the Sonic-3.5 voice model across all bots. The upgrade brings sharper intonation on emphasized words, lower time-to-first-byte for new utterances, and more natural pacing on longer responses.

No action needed. Existing bots pick up the new model automatically. Voice names, IDs, and the cloning pipeline are unchanged.

Bulk-call campaigns can now rotate across a pool of phone numbers instead of pinning to a single sender. Configure the pool once on the campaign, set a health threshold per number, and the dispatcher picks the healthiest number for each call.

What you get:

• Number pool per campaign. Add any active numbers in your org. The campaign auto-assigns from the pool at call time.
• Carrier health threshold. When a number's health score drops below the configured floor (rejection rates, busy patterns, carrier flags), the dispatcher routes new calls to other healthy numbers.
• Low-health alerts. Email notifications when a number falls below threshold so your team can investigate or pause.
• Automatic deactivation handling. If a number gets deactivated mid-campaign, the campaign rebalances to the remaining pool without manual intervention.

Available on every bulk-call campaign starting today. Open any campaign and configure the number pool from the campaign settings.

✨ What’s New: Navigation & AI Chat

• Command + K Search: A universal search bar that lets you jump to Bulk Calls, Phone Numbers, or Docs instantly.
• Assistant Search: Find any pre-built assistant without scrolling using the universal search.
• AI Chat: A chat interface designed to teach you how to build elite agents using platform-standard best practices.

 Feature Details

• Notification tone
  
  
• Phone number banner

Voice agents are only valuable when they can take real calls.

With OmniDimension Telephony, you can now purchase and use US (+1), and Indian (+91) phone numbers directly in OmniDimension [<2min], and, connect each number to a specific agent.

What you can do:

• Purchase numbers and concurrency (channels)

• Attach numbers to agents inside your dashboard

• Manage expiry and renewals without surprises

Buy phone numbers here:  https://omnidim.io/phone-numbers

Feature Details

• Numbers shop documentation
  
  
• Number purchase banner
  
  
• Agent Management APIs for IVR
  
  
• Improved Asterisk templates in Omnidimension
  
  
• Responsive UI fixes (tabs)

 Feature Details

• UI fixes on Phone Numbers page
  
  
• Bug fixes for number purchase and release flows

Feature Details

• Phone number purchase