Zum Inhalt springen

Vapi Release Notes

12 Einträge aus 1 Quelle. Zuletzt aktualisiert:

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: Freigabemodi in Composer, durchsuchbare Eval-Historie

Composer bietet pro Konversation wählbare Freigabemodi, die Eval-Laufhistorie ist durchsuchbar, mit Paginierung und per URL teilbar, und Call Logs sowie Voice-Fallback wurden verbessert und korrigiert.

What's New: Week of September 28, 2026

  1. Composer Per-Conversation Approval Modes: Composer now lets you choose when each conversation pauses for approval: ask before all changes, allow creates but ask before updates, deletes, runs, or publishing, bypass approvals entirely, or set custom permissions by resource and action.

  2. Eval Run History: You can now browse your complete Eval run history with search and pagination, and share a specific run result by URL.

  3. Fixes and Improvements:

    • Call logs: The logs table now shows clearer, deterministic error messages for 4xx responses instead of retrying unnecessarily, and the date picker now lets you select the full retention window.
    • Voice fallback: xAI terminal voice failures raised before a call starts are now delivered, the fallback voice no longer greets out of turn, and replacing a voice no longer consumes a second fallback.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: serviceTier „fast“ für GPT-5.6 und GPT-5.5, Provider-Request-IDs

Für OpenAI GPT-5.6 und GPT-5.5 lässt sich serviceTier auf fast setzen, um die schnellere Verarbeitungsstufe für latenzkritische Anrufe anzufordern, und bei Bring Your Own Key (BYOK) enthalten Call Logs nun die Request-IDs der Provider für Language Model, Transcriber und Voice.

What's New: Week of September 21, 2026

  1. Faster OpenAI Responses with Service Tiers: You can now set serviceTier to fast on your OpenAI GPT-5.6 and GPT-5.5 models to request OpenAI's faster processing tier for latency-sensitive calls.

  2. Provider Request IDs for Bring Your Own Key (BYOK): For BYOK models, call logs now include the provider's request ID for your language model, transcriber, and voice providers, so you can share it with the provider when getting support.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: Fallback-Modell für Anthropic Bedrock, Simulationen in Composer

Für Anthropic Bedrock kann im Assistant-Editor ein Fallback-Modell desselben Providers gesetzt werden, und Composer kann Simulationen samt Szenarien, Tester-Persönlichkeiten und Suites erstellen und ausführen.

What's New: Week of September 14, 2026

  1. Anthropic Bedrock Fallback Models: You can now set a same-provider fallback model for Anthropic Bedrock in the assistant editor.

  2. Simulations in Composer: Composer can now create and run simulations for you, including scenarios, tester personalities, and suites, so you can test your assistant right from the conversation.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: Chat testet ungespeicherte Konfiguration, Suites duplizierbar

Der Chat im Assistant-Editor testet jetzt die aktuelle ungespeicherte Konfiguration, Simulation Suites lassen sich duplizieren, und einen Tag vor der Abschaltung eines genutzten Modells kommt eine E-Mail.

What's New: Week of September 7, 2026

  1. Chat Tests Your Draft: In the assistant editor, Chat now runs your current unsaved configuration instead of the published version, so you can test your edits before publishing.

  2. Duplicate Simulation Suites: You can now duplicate an existing simulation suite.

  3. Model Retirement Notices: You now receive an email the day before a model your assistant uses is retired.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: Outbound-Kampagnen, Prompt Editor und benannte Transfer-Ziele

Im Dashboard lassen sich nun Outbound-Calling-Kampagnen erstellen und ausführen, außerdem gibt es einen Prompt Editor, Soniox-Endpoint-Einstellungen, benannte Transfer-Ziele und automatisch startende Videoaufzeichnung bei Web Calls (erfordert das neueste Web SDK).

What's New: Week of August 31, 2026

  1. Campaigns: You can now create and run outbound calling campaigns from the dashboard. Upload your contacts, pick an assistant, and track results as calls complete.

  2. Prompt Editor: The assistant editor now includes a prompt editor where you can generate a prompt from a short description, insert variables, and search within your prompt.

  3. Soniox Endpoint Controls: You can now tune the Soniox transcriber's endpoint sensitivity and endpoint latency adjustment level for finer control over turn taking.

  4. Named Transfer Destinations: You can now give each transfer destination a friendly name that shows in your call logs instead of a raw phone number.

  5. Auto-Start Video Recording: For web calls with video recording enabled, the recording now starts automatically when the caller joins, so your video recordings are captured reliably. This requires upgrading to the latest web SDK version.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: AssemblyAI Universal 3.5 Pro und Call-Log-Filter Websocket

AssemblyAI Universal 3.5 Pro steht als Transcriber zur Verfügung, und Call Logs lassen sich nach dem Call-Typ Websocket filtern.

What's New: Week of August 24, 2026

  1. AssemblyAI Universal 3.5 Pro: AssemblyAI's Universal 3.5 Pro model is now available as a transcriber for your assistants.

  2. Websocket Call Type Filter: You can now filter call logs by Websocket call type.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: Entwürfe und benannte Versionen, Webhook call.artifact.upload

Änderungen an Assistants und Tools gehen nicht mehr sofort live, sondern werden als Entwurf bearbeitet und als benannte Version veröffentlicht, außerdem gibt es den Webhook call.artifact.upload und die Region-Fixierung von GPT-5-Modellen auf Azure, z. B. gpt-5:eastus2.

What's New: Week of August 17, 2026

  1. Assistant and Tool Versioning: Edits no longer go live immediately. Work in a draft, publish when you're ready, and every publish creates a named version snapshot, so you can go back to your history to see what changed or roll back to a previous version as needed. You can also pin a specific tool version to an assistant.

  2. Call Artifact Upload Webhook: A new call.artifact.upload server message fires as each recording, packet capture, and log finishes uploading, and it includes the result for each, so you can process artifacts as they arrive instead of waiting for the end of call report.

  3. Azure Region Pinning for GPT-5 Models: You can now pin a GPT-5 model to a specific Azure region by adding the region to the model name after a colon (for example gpt-5:eastus2), which keeps requests in that region for lower latency and data residency. It uses Vapi's platform Azure credential, so no Bring Your Own Key (BYOK) is needed.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: Simulation Suites, Provider-und-Modell-Picker und SIP-fallbackPlan

Neu sind Simulation Suites zum Testen von Assistants und Squads, ein kombinierter Provider-und-Modell-Picker, Export in Chat- und Session-Logs, verbesserte SIP-Transfers mit fallbackPlan sowie zahlreiche Composer-Verbesserungen.

What's New: Week of August 10, 2026

  1. Simulations: You can now build simulation suites and run them against your assistants or squads to test and evaluate their behavior, with detailed results for each run.

  2. Unified Provider and Model Picker: The assistant editor now combines provider and model into a single picker for your LLM, voice, and transcriber.

  3. Chat and Session Log Export: You can now export selected rows from chat and session logs and use a bulk-actions bar, matching the call logs experience.

  4. SIP Transfer Improvements: Blind call transfers now support a configurable fallbackPlan, SIP verb, and dial timeout, and cold-transfer outcomes now appear in call logs.

  5. Composer Improvements: Composer is now more capable, reliable, and easier to use when building voice agents.

    • Upload files for it to reference, read current Vapi documentation directly, and follow new Vapi-built skills for building assistants, choosing models, managing tools, configuring phone numbers, and debugging calls.
    • Longer, multi-step tasks are far more reliable, with mid-task error recovery and progress that survives page refreshes and thread switches.
    • Connect Google Calendar from the conversation, clickable resource mentions with copyable IDs, and cleaner, collapsible activity timelines.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: Bedingte Structured Outputs und neue deutsche Deepgram-Aura-2-Stimmen

Structured Outputs können an Bedingungen geknüpft werden und werden nur bei erfüllter Bedingung erzeugt, und für Deepgram Aura-2 gibt es neue deutsche Stimmen.

What's New: Week of August 3, 2026

  1. Conditional Structured Outputs: You can now add conditions to a structured output so it only generates when the condition is met, and any skipped outputs are surfaced in the assistant preview, call logs, and sessions.

  2. Deepgram Aura-2 German Voices: New German voices are now available for Deepgram's Aura-2 voice.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: Claude Sonnet 5, GPT-5.5/5.6, Inworld TTS-2 und SIP-Warm-Transfer-Audio

Neue Modelle sind verfügbar (Anthropic Claude Sonnet 5, OpenAI GPT-5.5 und GPT-5.6 sowie Inworld TTS-2), und Warm Transfers bei SIP-Anrufen spielen nun Hold-Audio und einen Abschlusston (transferCompleteAudioUrl) ab.

What's New: Week of July 27, 2026

New Models

  1. Anthropic Claude Sonnet 5: Anthropic's Claude Sonnet 5 is now available as a model for assistants.

  2. OpenAI GPT-5.5 "Instant": OpenAI's GPT-5.5 (gpt-5.5 and chat-latest) is now available as a model for assistants.

  3. OpenAI GPT-5.6 Models: OpenAI's GPT-5.6 models (sol, terra, and luna) are now available for assistants.

  4. Inworld TTS-2: Inworld's TTS-2 voice is now available for assistants.

SIP Call Transfers

  1. SIP Warm Transfer Audio: Warm call transfers on SIP calls now play hold audio to the operator and a completion sound (transferCompleteAudioUrl) when the transfer connects.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: Model Presets und presigned Download-URLs im End of Call Report

Mit Model Presets lassen sich Transcriber-, LLM- und Voice-Modelle in einem Klick setzen und anhand von Latenz-, Kosten- und Qualitätsmetriken vergleichen, außerdem enthält der End of Call Report presigned Download-URLs für Aufnahmen und Logs und wird bei Zero Data Retention für transiente Assistants zuverlässig zugestellt.

What's New: Week of July 20, 2026

  1. Model Intelligence: You can now set your assistant's transcriber, llm, and voice models in one click with a Model Preset (choose between Balanced, High Intelligence, Ultra Fast, or Cost Saver), and see the latency, cost, and quality metrics for your chosen models so you can compare options and optimize with data.

  2. Recording Download URLs in the End of Call Report: The end of call report now includes short-lived presigned download URLs for your call recordings and logs, so you can download them directly.

  3. End of Call Reports with Zero Data Retention: End of call reports are now reliably delivered for transient assistants running under Zero Data Retention.

Originalquelle(öffnet in neuem Tab)Problem melden

Angaben zum Datum

Datum aus der Quelle.

Erstmals gesehen am .

Vapi

Vapi: Wiedergabegeschwindigkeit, Aufnahme-Downloads, VAD-Logs und HIPAA für xAI

Aufnahmeplayer haben eine Wiedergabegeschwindigkeits-Steuerung, alle Aufnahmetypen eines Calls sind herunterladbar, Call Logs zeigen VAD-Übergänge mit Latenzaufschlüsselung, und xAI ist durchgängig HIPAA-konform.

What's New: Week of July 13, 2026

  1. Playback Speed for Recordings: Call recording players now include a playback-speed control.

  2. Download Every Recording Type: You can now download every recording a call produced: mono, stereo, separate assistant and customer tracks, video, and packet capture.

  3. VAD Transitions in Call Logs: The call log now surfaces voice-activity-detection transitions with a per-phase latency breakdown.

  4. HIPAA Compliance: xAI is now HIPAA-compliant across its model, voice, and transcriber.

Originalquelle(öffnet in neuem Tab)Problem melden