> ## Documentation Index
> Fetch the complete documentation index at: https://docs.tavus.io/llms.txt
> Use this file to discover all available pages before exploring further.

# Changelog

<Update label="June 5" tags={["2026"]}>
  ## New Features

  * **Guardrails as first-class primitives:** Guardrails are now standalone resources that can be created, edited, and reused across multiple personas. Compose individual guardrails into sets via `guardrail_tags`. [Learn more](/sections/conversational-video-interface/guardrails)
  * **Choose your delivery channel:** Guardrail violations can now be delivered via [Interaction Events](/sections/conversational-video-interface/interactions-protocols/overview) (in-call `app-message`) or [Webhooks](/sections/webhooks-and-callbacks#guardrail-callbacks) (server-side `callback_url`) — or both.
  * **`guardrail_uuid` on violations:** Triggered events and callbacks now include `guardrail_uuid`, so you can identify exactly which individual guardrail was violated.
</Update>

<Update label="June 3" tags={["2026"]}>
  ## New Features

  * **Azure TTS support:** Azure is now available as a TTS engine, expanding language coverage for multilingual and localized voice output. [Learn more](/sections/conversational-video-interface/persona/tts)
</Update>

<Update label="May 22" tags={["2026"]}>
  ## New Features

  * **Individually addressable guardrails:** Guardrails are now first-class resources. Create, attach, edit, and delete each guardrail independently via the [Guardrails API](/api-reference/guardrails/create-guardrails). Bundle guardrails by tag and reference them on a persona via `guardrail_ids` or `guardrail_tags`. [Learn more](/sections/conversational-video-interface/guardrails)
  * **Node.js plugin support for LiveKit integration:** The LiveKit Agents integration now supports Node.js via the `@livekit/agents-plugin-tavus` plugin, in addition to Python. Install with `npm install @livekit/agents @livekit/agents-plugin-tavus`. [Learn more](/sections/integrations/livekit#node-js)
  * **Real-time event timestamps:** Every interaction event (`conversation.utterance`, `conversation.started_speaking`, etc.) now carries a `timestamp` field delivered in real time on the respective Interaction event. <a href="https://docs.tavus.io/sections/conversational-video-interface/interactions-protocols/overview" target="_blank">Learn more</a>
  * **Transcript utterance timestamps:** End-of-call transcripts now carry per-turn timing. The [`application.transcription_ready`](https://docs.tavus.io/sections/webhooks-and-callbacks#conversation-callbacks) webhook (and the same payload nested under `events` in the verbose [`GET /conversations/{id}?verbose=true`](https://docs.tavus.io/api-reference/conversations/get-conversation) response) now includes `timestamp` (Unix epoch float, seconds — same field name as live interaction events), `seconds_from_start`, `duration` (seconds, float — same field name as `conversation.stopped_speaking.duration`), and `inference_id` (on assistant turns) on each transcript entry.

  ## Enhancements

  * **New, and more detailed, error messages when a replica fails:** [Learn more](/sections/errors-and-status-details#replica-training-errors)
</Update>

<Update label="May 20" tags={["2026"]}>
  ## New Features

  * **AI Image Fixer API support:** [Create Replica](/api-reference/phoenix-replica-model/create-replica) now accepts an `auto_fix_training_image` property. Set it to `true` to use Tavus's AI Image Fixer to instantly fix any uploaded image to fit our requirements, eliminating the need for editing or recapturing photos. [Learn more](/sections/replica/train-with-an-image#ai-image-fixer)
</Update>

<Update label="May 12" tags={["2026"]}>
  ## New Features

  * **AI Image Fixer:** Instantly fix any uploaded image to fit our requirements, eliminating the need for editing or recapturing photos.
</Update>

<Update label="May 11" tags={["2026"]}>
  ## New Features

  * **Frame Checker (Video to Replica):** Get instant feedback, before recording, on whether your camera setup meets our requirements.

  ## Enhancements

  * **Less footage required (Video to Replica):** We now only require 1 minute of video, down from the 2 minutes previously needed.
  * **Simpler in-portal recording flow (Video to Replica):** A streamlined recording experience in the portal to help you capture high-quality recordings.
</Update>

<Update label="May 6" tags={["2026"]}>
  ## New Features

  * **Image to Replica:** Build a replica from a single still. Drop in a photo, illustration, or brand mascot.
</Update>

<Update label="May 8" tags={["2026"]}>
  ## New Features

  * **Voice Activity Detector improvements:** Resulting in a smoother conversational experience in noisy environments. This is automatically rolled out to all users.
  * **Expanded Recording Storage Support:** Conversation recordings can now be delivered to Google Cloud Storage (GCP) and Azure Blob Storage, in addition to AWS S3. <a href="https://docs.tavus.io/sections/conversational-video-interface/quickstart/conversation-recordings#set-up-your-storage" target="_blank">Learn more</a>

  ## Enhancements

  * **Tavus Components Library Updates:** Improved audio-video sync, plus new chat components and closed captions with streaming support. <a href="https://docs.tavus.io/sections/conversational-video-interface/component-library/components" target="_blank">Learn more</a>
</Update>

<Update label="May 1" tags={["2026"]}>
  ## New Features

  * **Wake Phrase:** Personas can now stay silent until they hear a specific phrase, similar to how voice assistants like Siri or Alexa work. Configure it via the `wake_phrase` parameter in the Conversational Flow layer. The persona still hears everything that is said and responds with full conversation history once the wake phrase is detected. <a href="https://docs.tavus.io/sections/conversational-video-interface/persona/conversational-flow#5-wake_phrase" target="_blank">Learn more</a>
</Update>

<Update label="April 30" tags={["2026"]}>
  ## New Features

  * **Idle Engagement:** Replicas can now proactively re-engage by speaking to the user after a period where the user is silent. Eagerness of this feature can be configured via the `idle_engagement` parameter in the Conversational Flow layer. <a href="https://docs.tavus.io/sections/conversational-video-interface/persona/conversational-flow#6-idle_engagement" target="_blank">Learn more</a>
</Update>

<Update label="April 29" tags={["2026"]}>
  ## New Features

  * **Speaking Events:** Two new events — `conversation.started_speaking` and `conversation.stopped_speaking` — fire for both the replica and the user with a `role` field (`"replica"` or `"user"`) identifying the speaker. `conversation.stopped_speaking` includes an `interrupted` boolean and a `duration` field (in seconds). <a href="https://docs.tavus.io/sections/event-schemas/conversation-started-stopped-speaking" target="_blank">Learn more</a>
  * **Conversation Diagnostics:** A new diagnostics surface for inspecting what happened in a conversation — including packet loss, network connection, FPS, and more — designed to make debugging significantly faster. Click on any conversation in the Developer Portal to access its diagnostics page.
</Update>

<Update label="April 17" tags={["2026"]}>
  ## Enhancements

  * **Non-Interruptible Custom Greetings:** Custom greetings now finish entirely before users can interrupt speech. Previously, participants could talk over a `custom_greeting`; now the replica completes the greeting before it begins listening. <a href="https://docs.tavus.io/sections/conversational-video-interface/faq#general-tavus-q-and-a:can-participants-interrupt-the-replicas-greeting" target="_blank">Learn more</a>
  * **Improved Turn-Taking Latency:** Significant TTS optimizations reduce turn-taking latency, resulting in faster and more natural back-and-forth during conversations.
</Update>

<Update label="April 10" tags={["2026"]}>
  ## New Features

  * **Streaming Utterance Event:** A new `conversation.utterance.streaming` event progressively reports what has been said during a conversation turn for both replica and user utterances. Use it to power closed captioning and build accurate transcripts — especially when a user interrupts the replica, since the streaming event reflects only the words actually spoken rather than the full LLM response. <a href="https://docs.tavus.io/sections/event-schemas/conversation-utterance-streaming" target="_blank">Learn more</a>
  * **Pronunciation Dictionaries:** Define custom pronunciation rules so your persona says brand names, technical terms, acronyms, and foreign words exactly right. Choose between simple alias substitution (e.g., "Tavus" → "TAH-vus") or precise IPA phonetic notation. Create a dictionary once and attach it to a persona via the TTS layer — any updates automatically propagate to all linked personas with zero extra latency at conversation time. <a href="https://docs.tavus.io/sections/conversational-video-interface/persona/pronunciation-dictionaries" target="_blank">Learn more</a>
</Update>

<Update label="April 7" tags={["2026"]}>
  ## New Features

  * **Voice Isolation:** Filter background noise from participant audio to improve conversation quality. Configure it via the `voice_isolation` parameter in the Conversational Flow layer. <a href="https://docs.tavus.io/sections/conversational-video-interface/persona/conversational-flow#4-voice_isolation" target="_blank">Learn more</a>
</Update>

<Update label="April 3" tags={["2026"]}>
  ## Changes

  * **Chat Interrupt History:** Personas now know when they have been interrupted. This allows the persona to pick back up where it left off, and also improves objectives adherence.
</Update>

<Update label="March 27" tags={["2026"]}>
  ## New Features

  * **Expanded ASR Model Selection:** You can now choose from five specialized speech-to-text engines via the `stt_engine` parameter. New models include `tavus-parakeet`, `tavus-soniox`, `tavus-whisper`, and `tavus-deepgram-medical`. Use `tavus-auto` to automatically route to the best model for each conversation. <a href="https://docs.tavus.io/sections/conversational-video-interface/persona/stt" target="_blank">Learn more</a>
</Update>

<Update label="March 19" tags={["2026"]}>
  ## Enhancements

  * **30% Faster Phoenix-4 Boot Time:** Phoenix-4 conversations now boot 30% faster, significantly reducing the time from conversation creation to readiness.

  ## Changes

  * **`conversation.replica_interrupted` Event Removed:** The `conversation.replica_interrupted` application message has been removed from interaction events. This event was deprecated in a previous backend update. Use `conversation.replica.stopped_speaking` with the `interrupted: true` property to detect interruptions instead.
  * **`duration` and `interrupted` Fields on Replica Stopped Speaking:** The `conversation.replica.stopped_speaking` event now includes a `duration` field (how long the replica spoke in seconds) and an `interrupted` field (`true`/`false`) indicating whether the replica was interrupted by the user. <a href="https://docs.tavus.io/sections/event-schemas/conversation-replica-started-stopped-speaking" target="_blank">Learn more</a>
</Update>

<Update label="March 17" tags={["2026"]}>
  ## New Features

  * **Event Ordering and Turn Tracking:** All server-broadcasted interaction events now include `seq` and `turn_idx` fields. `seq` is a globally monotonic sequence number for ordering events that may arrive out of order, and `turn_idx` groups related events from the same conversation turn. <a href="https://docs.tavus.io/sections/conversational-video-interface/interactions-protocols/overview#event-ordering-and-turn-tracking" target="_blank">Learn more</a>
</Update>

<Update label="March 11" tags={["2026"]}>
  ## Enhancements

  * **30% Faster Phoenix-4 Boot Time:** Phoenix-4 conversations now boot 30% faster, significantly reducing the time from conversation creation to readiness.
</Update>

<Update label="March 5" tags={["2026"]}>
  ## Enhancements

  * **EU ElevenLabs BYOK Support:** Customers can now bring their own ElevenLabs API key from EU-region accounts.
</Update>

<Update label="February 19" tags={["2026"]}>
  ## Enhancements

  * **Improved Knowledge Base Retrieval:** Optimized underlying infrastructure to improve utterance to utterance response times, particularly when `rag_search_quality` is set to `quality`.
</Update>

<Update label="February 17" tags={["2026"]}>
  ## New Features

  * **Expanded Tavus-Hosted LLM Selection:** Added new Tavus-hosted LLM options including models from Gemini, Claude, and GPT families. `tavus-gpt-oss` is recommended as the default. Legacy models `tavus-gpt-4.1`, `tavus-gpt-4o`, and `tavus-gpt-4o-mini` are now deprecated. <a href="https://docs.tavus.io/sections/conversational-video-interface/persona/llm#tavus-hosted-models" target="_blank">Learn more →</a>
  * **Visual RAG:** CVI now supports visual retrieval-augmented generation. Upload custom image explanations that are matched and queried via vision embeddings, giving your persona richer visual context during conversations.
</Update>

<Update label="February 12" tags={["2026"]}>
  ## Changes

  * **Persona**`context`**Field Deprecated:**  The `context` field has been deprecated in favor of a unified `system_prompt` field. Existing `context` values have been automatically merged into system prompts. The API remains backward compatible, but we recommend using **only** `system_prompt` going forward.
</Update>

<Update label="February 11" tags={["2026"]}>
  ## New Features

  * **Raven-1 Perception Model:** Introduced Raven-1, a multimodal perception model with audio emotion analysis and enhanced visual awareness. Raven-1 captures user emotion from audio in real time (sub-100ms audio perception latency), enabling personas to respond with greater emotional intelligence. The model is now the default for all new personas. Enable it by setting `perception_model_name` in your persona configuration. <a href="https://www.tavus.io/post/raven-1-bringing-emotional-intelligence-to-artificial-intelligence" target="_blank">Learn more →</a>
  * **Private Rooms:** Require authentication to join conversations for enhanced security. When enabled, we return a JWT meeting token that users must include when entering the room. <a href="https://docs.tavus.io/sections/conversational-video-interface/conversation/customizations/private-rooms" target="_blank">Learn more</a>
</Update>

<Update label="February 9" tags={["2026"]}>
  ## Enhancements

  * **Upgraded Transcription Engine:** Upgraded transcription engine with 3x improvements in word error rates (WER).
</Update>

<Update label="January 17" tags={["2026"]}>
  ## New Features

  * **Website Crawling for Knowledge Base:** You can now enable link crawling when creating knowledge base documents. Configure crawl `depth` and `max_pages` to automatically discover and ingest content from linked pages. Additionally, existing crawled documents can now be recrawled to keep knowledge base content up to date.
</Update>

<Update label="January 8" tags={["2026"]}>
  ## Changes

  * **PlayHT TTS Removed:** PlayHT has been fully removed as a supported TTS engine. All personas previously using PlayHT should migrate to Cartesia or ElevenLabs.
</Update>

<Update label="December 18" tags={["2025"]}>
  ## New Features

  * **Hard Delete for Conversations:** Conversations can now be permanently deleted via the API using the `hard=true` query parameter. Use this for GDPR compliance or data cleanup workflows.

  ## Enhancements

  * **Default TTS Model Updated to Sonic-3:** The default text-to-speech model has been updated to Sonic-3 across all new personas, delivering improved voice quality and naturalness.
  * **LiveKit Connection Stability:** Extensive reliability improvements to the LiveKit-based transport layer, including fixes for connection timeouts, track publishing hangs, event loop starvation, and ping timeout issues.

  ## Changes

  * **Default LLM Migrated to `tavus-gpt-oss`:** The default LLM for all new personas is now `tavus-gpt-oss`. All remaining `tavus-llama-4` personas have been automatically migrated. Legacy Tavus-Llama model references have been removed.
</Update>

<Update label="November 27" tags={["2025"]}>
  ## New Features

  * **LLM Temperature & Top-P Parameters:** You can now configure `temperature` and `top_p` parameters for both Tavus-hosted LLMs and custom LLMs via the `extra_body` field in your persona's LLM configuration. <a href="https://docs.tavus.io/sections/conversational-video-interface/persona/llm#4-extra-body" target="_blank">Learn more →</a>

  ## Enhancements

  * **Text Echo Language Accuracy:** Text echoes now correctly use the input language for conversion, improving accuracy in multilingual conversations.
</Update>

<Update label="September 17" tags={["2025"]}>
  ## New Features

  * **Test Mode for Conversations:** You can now start conversations in test mode, where the replica does not join. Validate your setup, integrations, and conversational flows without incurring costs or using concurrency slots. Set `test_mode: true` when creating a conversation. <a href="https://docs.tavus.io/api-reference/conversations/create-conversation#body-test-mode" target="_blank">Learn more →</a>
</Update>

<Update label="September 7" tags={["2025"]}>
  ## Enhancements

  * **Fuzzy Search for Personas:** Search now supports fuzzy matching for personas, allowing users to find results based on partial matches of UUIDs or names.
</Update>

<Update label="August 15" tags={["2025"]}>
  ## New Features

  * **Memories:** CVI now remembers context across conversations. Every conversation builds on the last with full context and time/date awareness, enabling use cases like adaptive tutoring, mentorship, and recurring consultations. <a href="https://docs.tavus.io/sections/conversational-video-interface/memories" target="_blank">Learn more →</a>
  * **Knowledge Base (RAG):** Bring your own data to conversations instantly. Upload documents or links and get grounded answers with \~30ms retrieval latency. Power AI recruiters, support agents, travel guides, and more with domain-specific knowledge. <a href="https://docs.tavus.io/sections/conversational-video-interface/knowledge-base" target="_blank">Learn more →</a>
  * **Objectives & Guardrails:** Define clear goals, branching logic, and measurable outcomes for your personas while keeping conversations safe, compliant, and on-brand. Ideal for complex workflows and regulated industries. <a href="https://docs.tavus.io/sections/conversational-video-interface/persona/objectives" target="_blank">Learn more →</a>
  * **Persona Builder:** A guided creation flow in the Developer Portal to shape AI personas with goals, behaviors, and style — then test or launch within minutes.
</Update>

<Update label="July 28" tags={["2025"]}>
  ## New Features

  * **Events Console:** A new events console in the Developer Portal lets you monitor everything happening during a conversation in real time — from message flows to system activity.
  * **Conversation Transcripts & Perception Analysis:** View full conversation details directly in the Developer Portal, including transcripts with speaker roles and perception analysis showing how your AI persona sees, hears, and responds.
</Update>

<Update label="July 25" tags={["2025"]}>
  ## New Features

  * **Persona Layer Controls:** Enable or disable layers like Sparrow directly within a Persona and adjust sensitivity settings in real time from the Developer Portal side panel.
  * **Persona Editing in Developer Portal:** We've added new editing capabilities to help you refine your Personas more efficiently. You can now update system prompt, context, and layers directly in our Developer Portal, plus duplicate existing Personas to quickly create variations or use them as starting points for new projects. Find these new features in your Persona Library at platform.tavus.io.
</Update>

<Update label="July 24" tags={["2025"]}>
  ## Enhancements

  * **Interaction Events Playground Improvements:** Major updates to the Interaction Events Playground including correct `properties.context` format and append vs overwrite toggle.
</Update>

<Update label="July 23" tags={["2025"]}>
  ## New Features

  * **Multilingual Settings in Developer Portal:** You can now specify the language of a conversation directly in the Developer Portal, including a new multilingual option for dynamic, real-world interactions.
</Update>

<Update label="July 22" tags={["2025"]}>
  ## New Features

  * **Llama 4 Support:** Your persona just got even smarter, thanks to Meta's Llama 4 model 🧠 You can start using Llama 4 by specifying `tavus-llama-4` for the LLM `model` value when creating a new persona or updating an existing one. Click <a href="https://docs.tavus.io/sections/conversational-video-interface/persona/llm#tavus-hosted-models" target="_blank">here</a>

    to learn more!
</Update>

<Update label="July 15" tags={["2025"]}>
  ## New Features

  * **React Component Library:** Developers can build with Tavus even faster now with our pre-defined components 🚀 Click <a href="https://docs.tavus.io/sections/conversational-video-interface/component-library/overview" target="_blank">here</a>

    to learn more!
</Update>

<Update label="June 27" tags={["2025"]}>
  ## New Features

  * **Multilingual Conversation Support:** CVI now supports dynamic multilingual conversations through automatic language detection. Set `properties.language` to "multilingual" and CVI will automatically detect the user's spoken language and respond in the same language using ASR technology.
  * **Audio-Only Mode:** CVI now supports audio-only conversations with advanced perception (powered by Raven) and intelligent turn-taking (powered by Sparrow-1). Set `audio_only=true` in your create conversation request to enable streamlined voice-first interactions.
</Update>

<Update label="June 20" tags={["2025"]}>
  ## Enhancements

  * **Fixed CVI responsiveness issue:** Resolved an issue where CVI would occasionally ignore very brief user utterances. All user inputs, regardless of length, now receive consistent responses.
  * **Expanded tavus-llama-4 context window:** Increased maximum context window to 32,000 tokens. For optimal performance and response times, we recommend staying under 25,000 tokens.
</Update>

<Update label="June 3" tags={["2025"]}>
  ## Enhancements

  * Reduced conversation boot time by 58% (p50).
</Update>

<Update label="May 28" tags={["2025"]}>
  ## Changes

  * Added a new recording requirement to <a href="/sections/replica/train-with-a-video" target="_blank">Training from a video</a>

    : Start the talking segment with a big smile.

  ## Enhancements

  * Added <a href="/sections/event-schemas/conversation-echo" target="_blank">echo</a>

    and <a href="/sections/event-schemas/conversation-respond" target="_blank">respond</a>

    events to conversational context.
</Update>

<Update label="May 17" tags={["2025"]}>
  ## Enhancements

  * **Major Phoenix 3 Enhancements for CVI**:
    * Increased frame rate from 27fps to 32fps, significantly boosting smoothness.
    * Reduced Phoenix step's warm boot time by 60% (from 5s to 2s).
    * Lipsync accuracy improved by \~22% based on AVSR metric.
    * Resolved blurriness and choppiness at conversation start.
    * Enhanced listening mode with more natural micro expressions (eyebrow movements, subtle gestures).
    * Greenscreen mode speed boosted by an additional \~1.5fps.
  * **Enhanced CVI Audio Quality**: Audio clicks significantly attenuated, providing clearer conversational audio.
  * **Phoenix 3 Visual Artifacts Fix**: Resolved visual artifacts in 4K videos on Apple devices, eliminating black spot artifacts in thumbnails.
</Update>

<Update label="May 9" tags={["2025"]}>
  ## New Features

  * Launched <a href="https://www.tavus.io/post/building-real-time-ai-video-agents-with-livekit-and-tavus" target="_blank">LiveKit Integration</a>

    : With Tavus video agents now integrated into LiveKit, you can add humanlike video responses to your voice agents in seconds.
  * <a href="https://docs.tavus.io/api-reference/personas/patch-persona" target="_blank">Persona API</a>

    : Enabled patch updates to personas.

  ## Enhancements

  * Resolved TTS (Cartesia) stability issues and addressed hallucination.
  * **Phoenix 3 Improvements**:
    * Fixed blinking/jumping issues and black spots in videos.
    * FPS optimization to resolve static and audio crackling.
</Update>

<Update label="April" tags={["2024"]}>
  ## Enhancements

  * **Replica API**:
    * Enhanced Error Messaging for Training Videos.
    * Optimized Auto QA for Training Videos.
</Update>