Skip to navigation

New Features & Improvements

  1. New speech vendors — speech-to-text: sarvam, smallestai, gnani, meta, saidwire, azureopenai and inworld. Text-to-speech: sarvam, smallestai, gnani, minimax, neuphonic, asyncai, respeecher, camb, hume, bland and soniox with streaming TTS, and vakyam, simplismart, groq, speechmatics, azureopenai, kokoro and spitch for say. Each has speech credentials in the portal and API. See TTS vendor settings.
  2. KugelAudio speech-to-text — kugelaudio, already a text-to-speech and noise isolation vendor, now provides streaming speech-to-text (kugel-ear-1). See KugelAudio.
  3. Soniox speech-to-text v6 — Soniox credentials and recognizers can use the stt-rt-v6 model.
  4. Speech-to-speech vendors in Agent Studio — Agent Studio adds Google Gemini (with a Gemini API key or a Vertex AI service account), xAI, Qwen, AssemblyAI, Azure VoiceLive, Azure OpenAI and Amazon Nova as realtime voice agents. These vendors were already available in the llm verb; each now has an account credential in the portal (LLM Services, Speech-to-speech tab) and a key test that does not open a billable session. These agents report turns to Agent Studio like the existing ones. See Realtime voice API.
  5. Sign in with GitHub or Google — a portal user can be created or edited to sign in with GitHub or Google instead of a password. The user receives an invitation email and is linked to their account on first sign-in by verified email address. Switching a user to GitHub or Google removes their password; switching back requires a new initial password, which they must change at first sign-in. GET /Users reports each user’s provider and whether the invitation is still pending.
  6. OpenAI Realtime errors — an OpenAI Realtime agent now reports an engine_error turn event when the vendor returns an error, a response fails, or the connection fails or drops, and the verb’s actionHook carries an error field.
  7. Speech-to-speech close reason — when a speech-to-speech vendor closes the connection, the disconnect now includes the vendor’s close code and reason.
  8. API version — GET /v1/Version returns the api-server release.
  9. Portal — call traces mark Dialogflow CES turns that the agent started without caller speech, and the hosted portal has the service provider switcher back for admins.

Changes in Behavior

  1. Turn detection by the speech vendor — with the agent verb, speech-to-text vendors that do their own turn detection (deepgramflux, assemblyai, speechmaticsagent, zoom, gradium, mistral, paxa and others) now end the caller’s turn on the vendor’s end-of-turn signal only, not also on the final transcript, which could run the turn twice. An end of turn with no words is ignored.
  2. Azure OpenAI realtime endpoint — Azure OpenAI realtime agents connect to the GA endpoint (openai/v1/realtime?model=) instead of the preview api-version and deployment endpoint.
  3. Gemini tool calls — Gemini realtime agents send one toolHook per tool call, carrying the vendor’s call id, instead of one per batch.
  4. Tenant checks — changing a SIP gateway’s carrier with PUT /SipGateways now returns 400. Assigning a phone number to an account outside your scope returns 403, and assigning it to an unknown account or application, or to an application of a different account, returns 400.
  5. Credential validation — a speech credential request missing a required field returns 400 instead of 500. LLM credentials that have never been tested now show as untested rather than failed, and are tested again when the portal lists them.

Removals

  1. Streaming TTS for Speechify, Giggy and Mistral — these vendors no longer support streaming text-to-speech (the agent verb, or say with stream: true). say without streaming is unaffected.

Bug Fixes

  • Hold on anchored-media dials — when a dial anchors media, the caller’s hold and resume are now answered by jambonz and passed to the onHoldHook.
  • Azure speech-to-text socket leak — speech-to-text vendor connections are now closed when recognition stops. Azure connections were left open, which over time caused gaps of up to 1.3 seconds in text-to-speech.
  • ElevenLabs data residency — ElevenLabs speech-to-text, and its credential test, use the host set by the credential’s data residency.
  • Deepgram Flux — Flux uses the recognizer’s model and language.
  • Speech-to-text connection failures — alerts and errors now give the vendor’s reason.
  • AssemblyAI agents — an AssemblyAI agent can be asked to reply mid-call.
  • VAD barge-in — when VAD barge-in stopped the agent for a sound the speech-to-text vendor never transcribed, the agent now resumes instead of staying silent.
  • OpenAI speech-to-text — OpenAI’s speech-stopped event now reaches the agent.
  • Dialogflow CES traces — traces of long CES calls are no longer dropped.
  • License changes — calls are no longer refused while a license change is being applied.
  • Recent calls — the list no longer empties after the date filter changes.

Component Versions

Component11.1.1111.1.12
jambonz11.1.1111.1.12
mediajam0.5.200.5.21
drachtio10.1.1210.1.12
rtpengine14.1.1.8-jambonz1414.1.1.8-jambonz14
jambonz OSS (sbc-*)0.9.150.9.15
upload-recordings1.9.31.9.3
monitoring-agent10.2.210.2.2
pcap-server1.0.51.0.5
heplify-server1.0.51.0.5

Availability

jambonz.cloud - November 7, 2026

self-hosted Debian packages - available now: install jambonz-mini 11.1.12 from the jambonz apt repository (Debian 12 bookworm), amd64 and arm64.

self-hosted on AWS - available now: CloudFormation templates and public AMIs for all nine server types in 30 regions, amd64 and arm64.

New Features & Improvements

  1. Saidwire speech-to-text — saidwire is available as a realtime STT vendor, with speech credentials in the portal and API. Saidwire follows Deepgram’s streaming protocol and is handled like deepgram: interim results, a final transcript per utterance, and jambonz’s own turn detection in the agent verb. It is English only and mono, and reports a fixed confidence of 0.95. saidwireOptions sets punctuate, smartFormatting, endpointing, utteranceEndMs and vadEvents. See recognizer.

New Features & Improvements

  1. Sarvam AI text-to-speech — sarvam is available as a TTS vendor, with streaming TTS and speech credentials in the portal and API. The Bulbul models bulbul:v4-flash (default), bulbul:v3 and bulbul:v2 cover eleven Indian languages, and the portal lists the bulbul:v4-flash personas by language. Options are pace, pitch, loudness, temperature, enable_preprocessing and dict_id, plus buffering controls for streaming. See Sarvam AI.
  2. Sarvam AI speech-to-text — the same sarvam credential serves realtime speech recognition with saaras:v3-realtime (default) or saaras:v4, in 23 Indian languages or with automatic language detection. Sarvam signals the start and end of speech and returns a final transcript after each pause. sarvamOptions sets the stream type, the transcript mode (including translation to English), the voice activity settings, a prompt and, for saaras:v4, key terms. See recognizer.

New Features & Improvements

  1. Smallest AI text-to-speech — smallestai is available as a TTS vendor, with streaming TTS and speech credentials in the portal and API. The Lightning models lightning_v3.1_pro (default), lightning_v3.1_pro_10_26 and lightning_v3.1 are supported, and the portal lists the Pro voices by language. Options are speed, language, number_pronunciation_language, date_format, math_notation, pronunciation_dicts and content_filter, plus buffering controls for streaming, and word-level timestamps make alignment events available for the voices that return them. See Smallest AI.
  2. Smallest AI speech-to-text — the same smallestai credential serves realtime speech recognition with Pulse (pulse, default, or the English-only pulse-2). hints are sent as Pulse’s keywords (up to 100 terms), and smallestaiOptions sets formatting, end-of-utterance timing, diarization, redaction, timestamps and detection features. See recognizer.

New Features & Improvements

  1. Azure OpenAI text-to-speech — azureopenai is available as a TTS vendor for one-shot say (no TTS streaming), with speech credentials in the portal and API. A credential holds the resource key, the resource endpoint and the deployment names. The default deployment is gpt-4o-mini-tts with the voice alloy, and options are speed, instructions and model_id. See Azure OpenAI.
  2. Azure OpenAI speech-to-text — the same azureopenai credential serves realtime speech recognition on a transcription deployment (gpt-live-transcribe by default, or gpt-realtime-whisper, gpt-4o-transcribe, gpt-4o-mini-transcribe). azureopenaiOptions mirrors openaiOptions: deployment, model family, prompt, languages, keywords, delay, noise reduction, turn detection and local voice activity settings. See recognizer.

New Features & Improvements

  1. Gnani text-to-speech — gnani is available as a TTS vendor for one-shot say (no TTS streaming), with speech credentials in the portal and API. The Timbre models timbre-v2.5 (default) and timbre-v3.0 cover ten Indian languages plus Hinglish, and the portal lists the voices by language. Options are model_id, language and speed. See Gnani.
  2. Gnani speech-to-text — the same gnani credential serves realtime speech recognition with prisma-v2.5 (default) or prisma-v3.0, in ten Indian languages or with automatic Indic detection. Gnani signals the start and end of speech and returns one final transcript per utterance. hints are sent as bias words or key terms, and gnaniOptions sets the transcript format, substitutions, context, idle timeout and voice activity settings. See recognizer.

New Features & Improvements

  1. Meta speech-to-text — meta is available as an STT vendor, with speech credentials in the portal and API, for realtime speech recognition with Meta’s muse-voice-transcribe-1.0. Meta signals the start of speech, sends interim transcripts and returns one final transcript per turn. The recognizer’s language is sent as a language bias, hints are sent as keywords, and metaOptions sets the mode, the partial transcript style, keywords, language bias and a session id. See recognizer.

New Features & Improvements

  1. MiniMax text-to-speech — minimax is available as a TTS vendor, with streaming TTS and speech credentials in the portal and API. The Speech models speech-2.8-turbo (default) through speech-01-hd are available, and the portal lists MiniMax’s system voices by language. Options include speed, volume, pitch, emotion, text normalization, language boost, voice effects, pronunciation dictionaries and voice mixing. Streaming sends text over a live connection and cancels on barge-in. See MiniMax.

New Features & Improvements

  1. Neuphonic text-to-speech — neuphonic is available as a TTS vendor, with streaming TTS and speech credentials in the portal and API. The portal lists Neuphonic’s stock voices by language, and account voices can be used by their voice id. Options are language, speed, temperature, ipa_key and, for streaming, timestamps. Streaming returns word-level timestamps, so alignment events are available. See Neuphonic.

New Features & Improvements

  1. Camb.ai text-to-speech — camb is available as a TTS vendor, with streaming TTS and speech credentials in the portal and API. Models are mars-flash (default), mars-pro, mars-instruct and the mars-8.1 betas; voices are numeric ids, and cloned voices work by their id. Options include speaking_rate, enhancement switches, inference_steps and the mars-instruct user_instructions. Streaming returns word-level timestamps, so alignment events are available. See Camb.ai.

New Features & Improvements

  1. Async text-to-speech — asyncai is available as a TTS vendor, with streaming TTS and speech credentials in the portal and API. Models are async_flash_v1.5 (default), async_flash_v1.0 and async_pro_v1.0; voices are Async voice ids. Options are model_id, language, speed_control and stability. Streaming does not return word timestamps, so alignment events are not available. See Async.

New Features & Improvements

  1. Respeecher text-to-speech — respeecher is available as a TTS vendor, with streaming TTS and speech credentials in the portal and API. Models are en-rt (default), ua-rt, ar-rt, de-rt, pt-br-rt and sv-rt, picked from the language unless model_id is set. Options are model_id and the sampling controls seed, temperature, top_k, top_p, min_p, presence_penalty, repetition_penalty and frequency_penalty. Streaming does not return word timestamps, so alignment events are not available. See Respeecher.

New Features & Improvements

  1. Bland AI text-to-speech — bland is available as a TTS vendor, with streaming TTS and speech credentials in the portal and API. The voice is a Bland voice UUID, which also selects the model; Karen is the default en-US voice. Options are expressiveness and stability. Streaming does not return word timestamps, so alignment events are not available. See Bland AI.

New Features & Improvements

  1. Simplismart text-to-speech — simplismart is available as a TTS vendor for one-shot say (no TTS streaming), with speech credentials in the portal and API. Models are Orpheus 3B (default), Veena for Hindi and Qwen 3 TTS on a dedicated deployment, set with the credential’s API host. Options are model_id, temperature, top_p, repetition_penalty and max_tokens, plus language and leading_silence for Qwen. See Simplismart.

New Features & Improvements

  1. Spitch text-to-speech — spitch is available as a TTS vendor for one-shot say (no TTS streaming), with speech credentials in the portal and API. Voices cover English, Yoruba, Hausa, Igbo, Amharic and Nigerian Pidgin; lina is the default en-US voice. Options are speed and api_uri. See Spitch.
Kokoro

jambonz Commercial (next release)