Audio

An Audio question asks candidates to respond by speaking. They can either record a voice response directly in the browser using their microphone, or upload an audio file. Audio responses are evaluated using AI or reviewed manually with a rubric.

Writing the Question

Write your question prompt using the content editor. The prompt should clearly describe what the candidate should say — the topic, the format (e.g. introduce yourself, describe the image, answer the question), and any constraints (e.g. speak for 1–2 minutes). See Content Editor and Layout.

Recording vs Upload

• Browser recording — candidates click a record button, speak into their microphone, and submit the recording. No software required. The browser requests microphone permission.
• File upload — candidates upload a pre-recorded audio file. Useful for assessments where candidates are given preparation time outside the platform.

AI Evaluation

Enable AI evaluation to automatically score audio responses. The AI first transcribes the spoken content, then evaluates the transcription against your grading prompt. The result includes:

• A score (success rate between -100% and 100%)
• AI feedback explaining the evaluation
• A transcription of the candidate’s spoken response

Write a grading prompt that describes what a good spoken response looks like — key content to include, language quality criteria, fluency requirements, etc. See AI Evaluation.

Rubric Scoring

Attach a rubric to evaluate audio responses against structured criteria — for example: pronunciation, vocabulary, grammar, content accuracy. Reviewers listen to the recording and score each criterion. See Rubric Scoring.

Common Features

Audio questions also support: Scoring — Multipliers & Dimensions, Hint, Answer Explanations, Folder & Tags, Preview & Scoring Simulation, and three save modes.