new·The score now tells you which way it movedA brain's exam only ever grows: its own material writes questions, and so does every question a real caller asked and did not get answered. The score is a percentage over that growing set, so a brain that learned more could post a smaller number — and this week three did. One of them answered two MORE questions than the week before and showed eighteen points less. Printed as a single percentage, that reads as decline to a reader and as punishment to anyone who contributes material.all news →
mozg.beta
Sign in

AI SDK · Providers · all subjects

elevenlabs/options

4 notes, read out of this brain and free to use. Each one was extracted from a source and is re-checked against its exam.

ElevenLabs speech model voice setting

When generating speech with ElevenLabs, the voice argument accepts a voice ID from the ElevenLabs Voice Library. Voice IDs can be found by selecting a voice in the library and copying its ID.

ElevenLabs speech provider options

The following provider-specific options are available for ElevenLabs speech generation: languageCode (string or null, ISO 639-1 code to enforce language, supported only for Turbo v2.5 and Flash v2.5); voiceSettings (object or null with properties: stability (double, 0-1 range, lower values increase emotional range, higher values create monotonous voice), useSpeakerBoost (boolean, boosts similarity to original speaker), similarityBoost (double, controls adherence to original voice), style (double, amplifies speaker style)); pronunciationDictionaryLocators (array of objects or null, up to 3 per request, each with pronunciationDictionaryId (string, required) and versionId (string or null, optional, uses latest if not provided)); seed (integer or null, 0-4294967295 for deterministic sampling); previousText (string or null, text before current request); nextText (string or null, text after current request); previousRequestIds (array of strings or null, max 3 IDs); nextRequestIds (array of strings or null, max 3 IDs); applyTextNormalization (enum: 'auto' (default), 'on', 'off', restricted to Enterprise plans for Turbo v2.5 and Flash v2.5); applyLanguageTextNormalization (boolean, defaults to false, helps pronunciation in supported languages like Japanese); enableLogging (boolean, defaults to account-level setting).

ElevenLabs transcription provider options

The following provider-specific options are available for ElevenLabs transcription: languageCode (string, ISO-639-1 or ISO-639-3 language code, defaults to null for automatic detection); tagAudioEvents (boolean, whether to tag audio events like (laughter), (footsteps), defaults to true); numSpeakers (integer, maximum speakers in file, max 32, defaults to null for maximum supported); timestampsGranularity (enum: 'none', 'word' (default), 'character'); diarize (boolean, whether to annotate which speaker is talking, defaults to true); fileFormat (enum: 'pcm_s16le_16' (16-bit PCM at 16kHz sample rate, single channel mono, little-endian byte order, lower latency) or 'other' (default for encoded waveforms)).

ElevenLabs streaming transcription provider options

The streaming provider options for ElevenLabs realtime transcription mirror ElevenLabs' realtime transcription settings: commitStrategy, enableLogging, filterBackgroundAudio, includeLanguageDetection, includeTimestamps, keyterms, minSilenceDurationMs, minSpeechDurationMs, noVerbatim, previousText, secondaryLanguages, vadSilenceThresholdSecs, and vadThreshold. When includeLanguageDetection is enabled, the provider requests the timestamp-bearing committed event internally but only returns timestamp segments when includeTimestamps is also enabled. filterBackgroundAudio cannot be combined with either includeLanguageDetection or includeTimestamps.

Give your agent this brain