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

assemblyai/capabilities

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.

AssemblyAI transcription result speaker diarization and audio-intelligence

The AI SDK's transcribe result exposes text, segments, language, and durationInSeconds. AssemblyAI's richer results including speaker diarization and audio-intelligence features are surfaced in two places: providerMetadata.assemblyai (structured results for enabled features including utterances for speaker-diarized segments when speakerLabels is set, entities, sentimentAnalysisResults, contentSafetyLabels, iabCategoriesResult, autoHighlightsResult) and response.body (complete raw AssemblyAI transcript response with any field not surfaced above).

AssemblyAI result timing units

Timings inside providerMetadata and response.body (e.g. utterances[].start) are in milliseconds matching the AssemblyAI API, whereas the top-level segments use seconds.

AssemblyAI speaker diarization example

Example of using speaker diarization with AssemblyAI: import { transcribe } from 'ai'; import { assemblyai } from '@ai-sdk/assemblyai'; import { readFile } from 'fs/promises'; const result = await transcribe({ model: assemblyai.transcription('universal-3-5-pro'), audio: await readFile('audio.mp3'), providerOptions: { assemblyai: { speakerLabels: true, entityDetection: true, }, }, }); const { utterances, entities } = result.providerMetadata?.assemblyai ?? {}; // utterances: [{ speaker: 'A', text: '…', start, end, … }, …] (start/end in ms)

AssemblyAI deprecated features

The following AssemblyAI features are deprecated by the API and not surfaced in providerMetadata (their output remains on the raw response.body if enabled): Summarization, Auto Chapters, and Custom Topics. Some features are language-gated (e.g. sentiment analysis is English-centric); see AssemblyAI documentation for per-language availability.

Give your agent this brain