Learn & Support

1. Create an account — Sign up with your email and a strong password (minimum 10 characters, at least one uppercase letter and one number).
2. Create a new recording — From the Recordings dashboard, click New to create a recording session.
3. Set your language — Choose your input language (the language being spoken) from the left panel. If you want translation, toggle it on and select an output language.
4. Choose audio source — Click the microphone icon to record from your mic, the monitor icon to capture system/tab audio, or the bot icon to send a meeting bot to join a video call (Pro/Max plans).
5. Start recording — Press the record button. Your transcript will appear in real time in the center panel.
6. Toggle features — Use the toggles in the left panel to enable Live Notes, Translation, Interpreter, and Captions as needed.
7. End session — When done, click the stop button. Your summary will be generated automatically. If you have auto-email enabled, a follow-up email will be sent to your recipients.

StellarCalls uses Deepgram Nova-3 for real-time speech-to-text transcription. As you speak (or as audio plays from a shared tab), words appear instantly in the transcript panel.

• Supports 17 languages as input — set your input language before recording to ensure accurate detection.
• Includes speaker diarization — the system detects different speakers and labels them (Speaker 1, Speaker 2, etc.).
• Transcript segments are timestamped. Click any segment to jump to that part of the audio playback after the session ends.
• Three view modes: Original, Translated (when translation is active), and Both (side-by-side).

When Live Notes is toggled on, AI automatically organizes your transcript into clean, structured notes in real time. Notes are sectioned by topic and update as the conversation progresses.

• Notes appear in the right panel while recording.
• They're generated in the input language of the session.
• Notes are saved automatically — they persist even if you leave the page and come back.
• Usage is tracked in minutes. Check your usage in the left panel or on the Usage page.

After a session ends, StellarCalls automatically generates a detailed, structured summary that includes:

• Summary — A 3–5 sentence overview of the session, including participants and conclusions
• Key Points — 5–10 bullet points covering the most important topics with specific details
• Decisions & Agreements — Any conclusions, positions, or agreements reached
• Action Items — Tasks, next steps, and follow-ups with responsible parties and deadlines
• Open Questions — Unresolved topics or items deferred for later

Summaries are generated in the session's input language and appear in the Summary tab. You can also manually trigger a summary by clicking the summary button.

Toggle Translation on and select an output language to see your transcript translated in real time.

• Supports translation between any combination of the 17 supported languages: English, Spanish, French, German, Chinese, Japanese, Arabic, Portuguese, Hindi, Korean, Farsi, Punjabi, Russian, Hebrew, Filipino, Indonesian, and Malay.
• Switch between Original, Translated, or Both transcript views.
• Captions (when enabled) show only the translated language while translation is active.
• You can change the output language mid-session — new translations will use the updated language.

Translation minutes are tracked separately from recording minutes. Check your usage on the Usage page.

The Live Interpreter goes beyond text translation — it speaks the translation aloud in real time using AI text-to-speech, like having a personal interpreter.

• Enable the Interpreter toggle and select your output language.
• The interpreter listens to the latest transcript and speaks the translation as the conversation happens.
• If the interpreter falls behind, it automatically skips ahead to keep up with the live conversation.
• The interpreter pauses automatically when you pause recording.

The Interview Copilot is an AI assistant that helps you during live conversations, interviews, or meetings.

• Auto-detect questions — Toggle this on and the AI will automatically detect when someone asks you a question, displaying it in the assistant panel.
• Suggested answers — Click on a detected question to get an AI-generated answer based on your uploaded documents and custom context.
• Upload documents — Add your resume, research papers, or reference materials. The copilot uses them to generate more relevant, specific answers.
• Custom context — Provide background about yourself or the conversation topic to improve answer quality.
• Quick actions — "What should I say next?", follow-up questions, fact-check, and recap buttons.
• Screenshot analysis — Share a screenshot for the AI to analyze and provide insights.
• Mini window — Pop the copilot out into a mini floating window so you can use it alongside your meeting app.

The Meeting Bot is a fourth audio source option that sends an AI bot to join your meeting on Zoom, Google Meet, or Microsoft Teams — no screen sharing or tab audio capture needed.

• Select Bot as your audio source, paste the meeting link, and press record. The bot joins the meeting and listens.
• The full transcript is delivered after the meeting ends — there is no live transcript during the call.
• If you pause the recording during the meeting, the paused portion is automatically excluded from the transcript.
• When the meeting ends, an AI summary is generated automatically from the transcript.
• While in bot mode, features like live notes, translation, interpreter, and captions are hidden since they don't apply.
• After the meeting ends, you can use AI Chat to ask questions about the transcript.
• If the recording is paused for more than 20 minutes, the bot will automatically disconnect.

Send a clean, professional follow-up email with your AI summary to colleagues or stakeholders directly from StellarCalls.

• Send from the summary panel — After a session ends and the summary is generated, click "Send Follow-up Email" in the Summary tab.
• Manage recipients inline — Click the Recipients button next to the send button to add or remove email addresses without leaving the recording view.
• Auto-email toggle — Enable "Auto-email after every meeting" in Settings to automatically send the summary to your saved recipients when any recording ends.
• Recipient management — Add up to 10 recipient email addresses in Settings > Email Follow-ups. Frequently used emails are saved automatically.
• The email includes a formatted version of your summary — it does not include the full transcript.

The Chat tab in the notes panel lets you ask questions about your recording in natural language.

• Ask about specific topics discussed, get clarifications, or request the AI to analyze parts of the conversation.
• Chat is context-aware — it has access to your full transcript.
• Chat messages count toward your monthly AI chat limit.

Share your transcript, notes, and summary with anyone via a public link.

• Click the Share button in the session controls to generate a shareable link.
• Anyone with the link can view the transcript, notes, and summary — no account required.
• You can revoke the link at any time from the same share dialog.
• Sharing is available on Basic plans and above.

Export your notes, transcript, and summary in multiple formats:

• PDF — Clean, formatted document ready to print or share.
• Word (.docx) — Editable document (Basic plan and above).
• LaTeX (.tex) — For academic use (Basic plan and above).
• Copy to clipboard — Quick copy with one click.

Free plan users get 1 export per month. Basic and above get unlimited exports.

Your recordings, transcripts, notes, and summaries all count toward your account storage limit.

View your current usage at any time on the Usage page. Each feature toggle in the session view also shows a compact usage counter (e.g. 12/300m) next to it.

StellarCalls is a web application. It runs in your browser — no download required. The following browsers are supported:

Full support (Chromium-based)

• Google Chrome (recommended)
• Microsoft Edge
• Brave
• Arc
• Opera

Limited support

• Firefox — Basic transcription works, but system audio capture is not supported. Some real-time features may behave inconsistently due to limited MediaRecorder and Web Audio API support.
• Safari — Microphone recording works on recent versions, but system audio capture and some real-time features are not available. Safari on iOS has additional constraints on background audio processing.

Audio input

StellarCalls supports four audio sources:

• Microphone — captures your voice. Ensure your browser has microphone permission enabled.
• System audio (tab audio) — captures audio from another tab (e.g. a Zoom or Google Meet call). This uses Chrome's screen/tab sharing API. When prompted, select the tab whose audio you want to capture.
• Both — captures from both your microphone and system audio simultaneously.
• Meeting Bot — sends an AI bot to join a Zoom, Google Meet, or Microsoft Teams meeting. No screen sharing required. Available on Pro and Max plans.

Other requirements

• Stable internet connection (audio is streamed to our servers for transcription)
• JavaScript must be enabled
• Works on desktop and mobile — but desktop is recommended for the full feature set

• Browser support — Chromium browsers only (Chrome, Edge, Brave, Arc, Opera). Firefox and Safari have limited compatibility.
• System audio — Only available on desktop Chromium browsers via tab sharing. macOS does not support full system audio capture natively.
• Concurrent recording — You can only record on one device at a time per account. Starting a recording on a second device will be blocked.
• Session length — Maximum session length depends on your plan (30 min for Free, up to 4 hours for Max).
• Device sessions — You can be logged in on up to 3 devices simultaneously. Logging in on a 4th device will require signing out of another.
• Offline mode — StellarCalls requires an internet connection. Transcription and AI features will not work offline.
• Transcription accuracy — Accuracy varies by language, audio quality, background noise, and speaker clarity. English generally has the highest accuracy.
• Interpreter latency — There may be a 2–5 second delay between speech and the interpreter's spoken translation due to processing time.