Official MurMur information
Voice Typing, Recorded-Audio Transcription or Translation?
Start with the result you need: text in a writing app, a record that stays close to your wording, an organized draft or an expression in another language. This task-based guide helps you decide whether MurMur's current features fit your workflow.
MurMur · Content reviewed:
Workflow comparison
| Your task | A suitable workflow | What to check |
|---|---|---|
| Write a thought in a chat, email or note | A voice keyboard, such as MurMur | Suitable for placing spoken content in supported text fields; first check app compatibility and keyboard permissions. |
| Stay close to the words you originally used | Direct transcription | Limits cleanup-related changes to phrasing; compare with the original material when the record must be verbatim. |
| Turn a spoken email or several tasks into a draft | Smart cleanup | Can help with repetition, punctuation and paragraphs; review for missing requirements or changes in meaning. |
| Keep Chinese and English terms in one sentence | A supported mixed-language recognition mode | Useful for bilingual writing; personal vocabulary can help with proper names, which still need checking. |
| Convert the whole message into another language | Translation supported by the current app | Requires a supported target language and internet access; mixed-language dictation is not itself translation. |
| Process an existing meeting or interview recording | A transcription tool matched to the recording requirements | Check file import, recording length, timestamps and speaker identification. Voice-input support alone does not establish these capabilities. |
| Write without an internet connection | A supported local mode with its model ready | Test it on the actual device first; cloud recognition, smart cleanup and translation require a connection. |
Start with writing or an existing recording
When replying to a message or drafting an email, a voice keyboard is about placing your current spoken thoughts in a text field. Check whether your usual apps allow third-party keyboards and whether the text reaches the intended place.
When the task starts with a meeting or interview file, list your requirements for file formats, recording length, timestamps and speaker identification. Then check the chosen tool's explicit feature documentation. Voice input alone does not establish suitability for those recording tasks.
Decide whether wording or structure matters most
Decide which parts may be changed. For specific phrasing, start with direct transcription and check it against the original. For a draft made from spoken ideas, try smart cleanup, then review paragraphs, lists and tone.
With either output, check numbers, dates, negations and corrections made while speaking. Neater formatting cannot establish that the content is accurate.
Distinguish mixed dictation from translation
To keep English terms inside a Chinese sentence, choose a mode that supports mixed Chinese-English dictation. To make the whole message readable in another language, check the translation feature and target languages available in your current app.
Both tasks need a review of names, numbers and specialist terms. Translated wording also needs to suit the reader's context. Check the privacy notice for cloud processing and connectivity requirements.
Evaluate a short example from your own work
Choose a message you often write, including a name, a date and a familiar work term. Try it in a note. Check whether the information is complete, the text reaches the correct place, the editing needed is acceptable and the mode's connectivity requirements fit your situation.
If you need cleanup or translation, use the same content to check whether requirements and meaning are retained. This helps you choose a workflow; it is not an accuracy or speed ranking.