AI Meeting Summaries
Control how AI generates meeting summaries with custom instructions, language settings, and regeneration.
What the AI Generates
Control how AI generates meeting summaries by setting company and personal instructions, choosing the transcription language, regenerating with free-text guidance, and rewriting selected passages.
For the full summary studio, sharing, and PDF workflow, see Meeting Summaries.
The AI processes the meeting transcript in chronological order together with the participant roster. It then produces a structured summary with a title, TL;DR, and context-specific sections.
Output Structure
| Element | What the AI produces |
|---|---|
| Title | A concise, concrete title for the meeting |
| TL;DR | A two-to-three sentence overview |
| Sections | Context-specific sections covering observations, identified issues or deviations, follow-ups and commitments, deadlines and next steps, and responsible people for action items |
The AI corrects obvious speech-to-text errors and uses the exact participant name spellings from the meeting roster.
Output Limits
| Constraint | Limit |
|---|---|
| Maximum sections | 4 H2 sections |
| Maximum bullets per section | 5 |
| Format | Markdown converted to HTML, with no JSON or filler commentary |
What You Control
Company custom instructions
Company custom instructions influence the emphasis and writing style of generated summaries. Users with the meeting settings permission can set, save, and reset these instructions.
Use company instructions to focus summaries on recurring priorities. For example, you can ask the AI to focus on safety observations or include cost estimates whenever participants mention them.
Personal host instructions
Personal host instructions apply to meetings you host. They take priority over company instructions when the two conflict.
Personal instructions support up to 4,000 characters. Use them for preferences that apply to your meetings without changing the default for the rest of the company.
Transcription language
Set the language at the company level, override it for your meetings, or choose it for an individual meeting.
| Level | What it controls |
|---|---|
| Company default | Sets the default language for all company meetings |
| Personal override | Chooses a language for your own meetings |
| Per-meeting | Selects a language when scheduling or hosting a meeting |
You can reset your personal language to the company default.
Regeneration
Regeneration accepts a free-text instruction and applies it as a higher-priority refinement layer to the summary. Ask for a different emphasis, a shorter version, a more focused presentation, or another language.
When you request a translation, the AI changes the full output language, including the title and headings.
AI selection-edit
Selection-edit rewrites a selected passage of the summary according to a free-text instruction. Use it to clarify wording, change emphasis, or adjust a specific passage without regenerating the entire summary.
Selection-edit has a 30-second timeout and uses separate metering from the main summary generation.
Manual editing and version history
You can edit the generated summary manually and review its version history. Manual corrections create new versions, so the original summary remains available in history.
Industry preset
Users who can manage the company profile choose one industry preset for the company from a list, or choose no preset. The presets are managed centrally by Vadio administrators, who write each preset's instructions and vocabulary list. Companies cannot view or edit the vocabulary list. Use company custom instructions to add your own guidance.
Plan Limits
Meeting summary generation, regeneration, and selection edits consume the AI summary token pool. When the pool is exhausted, these features are blocked until the quota resets or is increased.
| Plan | AI Summary Token Pool |
|---|---|
| Free | 0 |
| Trial | 250,000 |
| Start | 1,000,000 |
| Standard | 2,000,000 |
| Premium | 4,000,000 |
The allowance resets every month, also on yearly plans. Unused tokens do not roll over to the next cycle.
When the pool is used up, Summary Studio shows a notice titled "AI summary limit reached for this cycle". The notice says that new generation is paused, gives the date the allowance resets, and confirms that everything already generated stays usable. The guidance depends on who is looking:
- Users who can manage billing see a Chat with support button.
- Other users are told to contact their admin or owner to review the limit.
- Users in a chain company see a notice titled "AI limit reached for this cycle" instead. It says whether the shared chain pool or the company's own chain limit was reached, and that only Vadio can lift it.