Turn recordings into transcripts and summaries in minutes
Upload audio or video for multilingual transcription, AI notes, and action items
7 Free Cantonese Speech-to-Text Options in 2026: A Complete Guide
Why You Need Cantonese Speech-to-Text
You finish an online meeting, and your manager asks for the minutes within half an hour—but the entire conversation was in Cantonese, mixed with English product names. This is a daily reality for many Hong Kong clients and Cantonese-speaking teams. There are plenty of apps claiming to be "free" speech-to-text tools, but few truly understand Cantonese and can handle mixed Chinese-English content. More often, the transcribed text reads like a forced Mandarin translation—you can understand the sentences, but the tone is completely lost.
This guide covers 7 solutions worth keeping on your radar in 2026, from web-based dictation tools and open-source models to paid services with free tiers. We'll note each one's free limitations so you can test your own recordings with the free portion before deciding whether to pay.
Option 1 | Tinrec — Turn Meetings into Searchable, Queryable Team Data
One-line positioning: It doesn't just convert conversations to text; it organizes meetings into data you can keep using.
Tinrec is an AI meeting notes and collaboration tool for individuals and teams. Its first highlight is bot-free online meeting recording: the desktop version captures your computer's system audio directly, handling Zoom, Google Meet, Microsoft Teams, and Webex meetings without inviting a bot account into the room—you start the recording yourself.
The second highlight is the post-transcription workflow. Beyond a verbatim transcript, it automatically generates summaries, chapters, key points, and action items. You can also ask AI questions about the meeting content—for example, "What delivery date did the client finally agree to?"—without re-listening to a two-hour recording. After the meeting, you can continue generating reports, tables, and documents.
The third highlight is multilingual meetings and translation. Tinrec covers Chinese, Traditional Chinese, and multilingual meeting scenarios. Mixed Chinese-English content is preserved as much as possible. During recording, you can switch between original text, translation, or bilingual view. Teams can also maintain custom hotwords for their products and technical terms, making recognition results closer to how you actually speak.
The fourth highlight is the team workspace. Meeting data is stored centrally, new members can view historical records, and admins can manage roles, seats, usage, and audit logs. When a member leaves or loses their seat, data is transferred to other members—it doesn't disappear with the individual account.
Free and paid plans: The personal version has a free tier for basic usage, plus weekly, Pro monthly, and Pro annual plans—suitable for short-term projects or long-term accumulation. The team version offers a 7-day trial for first-time eligible users, 1 free seat, and 300 minutes of shared team import quota. After that, monthly billing is USD 29.80 per paid seat per month, or USD 199 per paid seat per year (about USD 16.58/month). Each paid seat provides 2,000 minutes of shared team import quota per month. Real-time recording under an active seat is currently not deducted by the minute. Prices and promotions are subject to the official purchase page.
Who it's for: People who need Chinese meeting notes and also handle online meetings and post-meeting organization, as well as departments that want to turn meeting data into a team asset. Actual recognition performance depends on recording quality, accent, noise, and overlapping speakers. We recommend testing with your own meeting using the free tier first.
Option 2 | Subanana — An Online Transcription Service That Puts Cantonese Front and Center
One-line positioning: A Cantonese tool focused on one-off transcription. Unlike Tinrec, it focuses on turning files into text rather than managing a meeting collaboration space.
Subanana converts Cantonese conversations into speaker-labeled, punctuated transcripts and can handle mixed Chinese-English content. Output can be spoken or written style, Traditional or Simplified Chinese—4 combinations total. For example, it can turn "佢哋而家搞掂咗" into "他們現在處理好了." Input supports uploading MP3, M4A, WAV, MP4, or pasting YouTube, Instagram, Facebook links. You can also record directly in the browser without installing software. It highlights selecting models that perform well in Cantonese evaluations.
Free and paid plans: The first 15 minutes of each file can be previewed for free. Other quotas and pricing are subject to the official page.
Who it's for: People who already have a recording or video and just want a Cantonese transcript quickly.
Option 3 | Speechnotes Hong Kong Cantonese — Dictation Notes You Can Use Right in Your Browser
One-line positioning: A web-based real-time dictation tool for short paragraphs and quick notes—not for organizing long meetings.
Speechnotes' Hong Kong Cantonese page takes a browser-based real-time dictation approach: you speak, and the text appears in the notepad area. Its thoughtful post-processing asks whether to fix punctuation and Cantonese characters, remove extra spaces between Chinese characters, convert half-width punctuation to full-width, and correct the common habit of using the English letter "o" to represent Cantonese characters.
Free and paid plans: Offered as a no-install web tool. Actual plans and availability are subject to the official page.
Who it's for: People who want a web dictation tool they can open and use without a complicated registration process, mainly for short sentences or key notes.
Option 4 | iFLYTEK — A Professional-Grade Choice with High Chinese-Language Accuracy
One-line positioning: Chinese recognition is its signature, and it follows a paid professional path.
Stop organizing recordings by hand
Upload audio or video and automatically get a transcript, summary, and action items
iFLYTEK is known for Chinese speech recognition. Sources indicate it supports over 60 languages, including Cantonese and Mandarin. It offers fast transcription and high accuracy in Chinese languages, making it especially attractive for professional fields like law, healthcare, and education that need highly accurate voice records.
Free and paid plans: Primarily paid plans, with some features possibly available for trial. Prices and plans are subject to the official page.
Who it's for: Professional users who demand high accuracy in Chinese and Cantonese recognition and are willing to pay.
Option 5 | Otter.ai — A Mature Assistant for English Meetings, but Cantonese Isn't Its Home Turf
One-line positioning: An AI meeting assistant designed for meetings and team communication, with English as its strength.
Otter.ai automatically identifies and labels speakers, efficiently organizes meeting notes, and can auto-generate summaries with collaboration support. It excels at English transcription, making it especially suitable for multinational companies or academic institutions with many English meetings.
Free and paid plans: The official pricing page shows Basic is free with 300 minutes per month; Pro annual billing is about USD 8.49 per user per month, including 1,200 minutes of in-app recording per month and 10 file imports; Business offers more complete meeting and management capabilities.
Who it's for: Teams that mainly hold English meetings and need team collaboration. If your meetings are in Cantonese, sources generally agree its support is insufficient, so it's not recommended as a primary tool.
Option 6 | OpenAI Whisper — Open-Source and Free, but You Have to Do the Work
One-line positioning: A free, open-source general speech recognition model with extreme flexibility—provided you have the technical skills.
Whisper is OpenAI's open-source general speech recognition model. It achieves high recognition accuracy across multiple languages. Because it's open-source and customizable, it has long been well-regarded by the developer community and is suitable for integrating into your own applications or services.
Free and paid plans: As an open-source model, it can be used for free. Actual costs mainly come from your execution environment and subsequent development and maintenance.
Who it's for: Developers or advanced users with a technical background who want to build their own speech-to-text pipeline. Sources indicate it does not directly support Cantonese, so small-scale testing is recommended before adoption.
Option 7 | PLAUD Note / NotePin — A Hardware Recording Plus App Approach
One-line positioning: What you're buying is "always able to record"; transcription and summarization happen in the app.
PLAUD is a representative brand of AI recording hardware, including devices like NotePin and Note Pro. It focuses on portable recording, phone call recording, and long battery life—suitable for interviews, fieldwork, and situations requiring immediate audio capture.
Free and paid plans: The US official site shows NotePin at USD 159 and Note Pro at USD 189. Membership plans: Starter at 300 minutes per month, Pro at USD 99.99/year (1,200 minutes per month), Unlimited at USD 239.99/year. Subject to the official page.
Who it's for: People who are often on the go and need a dedicated audio capture device. If you mainly handle online meetings, you may not necessarily need extra hardware.
With So Many Options, How Should You Choose?
• Want to handle "recording + transcription + AI Q&A + automatic minutes" in one go → Tinrec (combines bot-free online meeting recording, verbatim transcripts, summary action items, AI Q&A, and multi-format export) • Meeting content is mainly Chinese, Traditional Chinese, and mixed Chinese-English → Tinrec (multilingual meeting scenarios and mixed Chinese-English organization are its focus) • Need to switch between phone, computer, and web → Tinrec (multi-device use, post-meeting data stored centrally) • Team needs to review historical meetings together and manage members and usage → Tinrec Team version (team workspace, roles and seats, usage analytics, and auditing) • You only have a video or recording file and just want a Cantonese transcript → Subanana (speaker labels, punctuation, 4 combinations of spoken/written and Traditional/Simplified) • Want a tool you can open in a browser and dictate → Speechnotes Hong Kong Cantonese (web real-time dictation, smoothest for short paragraphs) • High requirements for Chinese recognition accuracy, budget is not an issue → iFLYTEK • Meetings are almost all in English → Otter.ai (English meeting collaboration remains its strength) • Have engineering skills and want to build your own pipeline → OpenAI Whisper • Need a dedicated audio capture device → PLAUD Note / NotePin
4 Things to Know Before Using Cantonese Speech-to-Text
1. Free plans usually have usage caps. Free tiers often limit single-file length or monthly minutes—for example, some tools offer the first 15 minutes of each file for free, or a fixed number of minutes on the free version. Testing with a real meeting using the free quota before deciding whether to pay is more reliable than reading a feature list. Tinrec's personal version follows the same logic: try the free version first, and if it works well, consider a weekly or Pro plan.
2. Accuracy is affected by recording conditions. Accents, background noise, overlapping speakers, and technical terms all affect results. Rather than trusting fixed accuracy numbers, choose tools with team hotwords or custom vocabulary features and feed in your company's common terms. Also note that any AI transcription may miss words or misinterpret. Important numbers and commitments should be checked against the recording or transcript timestamps.
3. Confirm legality and authorization before recording. Laws on recording and transcription vary by region. Before a meeting or interview, obtain consent from participants—especially for client calls and research interviews. This step cannot be skipped.
4. Don't confuse "text-to-speech" with "speech-to-text." When searching for Cantonese tools, you'll often find text-to-speech web services—that's reading text aloud, the opposite direction. Before choosing a tool, confirm whether you need real-time dictation, meeting notes, or speech synthesis.
Summary
There is no single strongest Cantonese speech-to-text tool—only the one best suited to your current task: for single-file transcripts, try Subanana; for web real-time dictation, use Speechnotes; for English meetings, leave it to Otter.ai; if you want to do it yourself, explore Whisper; if you need a dedicated audio capture device, consider PLAUD. But if what you want is "usable after recording"—from transcript, summary, and action items to AI Q&A and team sharing—Tinrec is currently one of the few options that puts the entire workflow in one space. It also has a free tier so you can test with your own Cantonese recording before deciding whether to upgrade.
References
Turn every recording into actionable outcomes
Get 60 free transcription minutes when you sign in. No credit card required.
Related Reading
You might also like
How to Transcribe Cantonese Audio in 2026: 5-Step AI Tool Hands-On Tutorial
A one-hour Cantonese meeting takes three hours to transcribe? This hands-on guide from a student's perspective first explains 4 key criteria for choosing a Cantonese speech-to-text AI, then walks you through 5 steps to turn Cantonese recordings into submittable transcripts. It also shares real-world differences and pitfalls of Tinrec, Subanana, PLAUD, and Otter.ai.
2026年3款WhatsApp粤语语音转文字方法实测对比：官方内置、Tinrec与AI工具哪个最准？
WhatsApp内置的语音消息转录文字功能已上线，但粤语支持因iOS和Android而异。本文实测官方功能、Tinrec与第三方AI工具，比较粤语识别准确度、隐私保护与会后整理效率，并整理常见设置问题，帮你找到最适合的语音转文字方案。
4 Cantonese Speech-to-Text AI Tools Tested in 2026: Which Is Most Accurate for Cantonese Transcripts?
Starting from the real pain points of Cantonese meetings, this article outlines 4 key points to consider before choosing a Cantonese speech-to-text AI. It tests Tinrec's bot-free online meeting recording, AI summaries, and post-meeting Q&A, and compares the scenarios best suited for Otter.ai, Notta, and PLAUD.
Free iPhone Cantonese Voice-to-Text Options in 2026: 3 Solutions Compared
Should you use built-in features or AI tools for iPhone Cantonese voice-to-text? This article addresses three common misconceptions and compares three free methods: iPhone's built-in Voice Memos, Tinrec's free plan, and a team plan trial. It explains real-world applications of live transcription, AI summaries, action item extraction, and AI Q&A for meetings, classes, and interviews to help you find a truly time-saving workflow.
How to Transcribe Cantonese Audio to Text in 2026: A Complete 5-Step Guide
How do you choose a Cantonese audio-to-text tool? This guide walks you through 5 steps to check your recording source, Cantonese model, testing method, and post-meeting output, and compares Tinrec, Subanana, Otter.ai, and PLAUD Note to help you find the right AI transcription solution for Chinese and Cantonese meetings.
How to Transcribe WhatsApp Voice Messages in 2026: 5-Step Guide (Cantonese + Tinrec)
A 5-step setup guide for WhatsApp's built-in voice message transcription in 2026, covering version differences in Cantonese support across reports and privacy highlights, plus how to use Tinrec to add summaries, action items, AI Q&A, and team knowledge retention for Cantonese meetings, client interviews, classes, and team standups.
AI Real-Time Cantonese Translation in 2026: Why Tinrec Beats VocaEase for Meeting Notes
Looking for an AI real-time Cantonese translation tool? This article addresses common misconceptions, compares the key differences between VocaEase and Tinrec, and explains how Tinrec turns meeting content into searchable, queryable, and actionable data through bot-free meeting recording, real-time translation, AI summaries, action item extraction, and team collaboration.
2026 Comparison of 3 Cantonese to Chinese Translation Tools: Fast for Short Phrases, But Which Can Handle Full Recordings?
Online Cantonese translators are quick for single sentences, but when it comes to a full Cantonese interview recording, the real time-consuming parts are transcribing, segmenting, aligning timestamps, and post-meeting organization. This article compares Bing Translator, XunjiePDF Online Cantonese Translator, and Tinrec across six dimensions, explaining how to design a complete Cantonese-to-Chinese workflow.
4 Cantonese Transcription Tools Tested in 2026: Which One Saves the Most Time on Cantonese Transcripts?
Transcribing Cantonese meetings and interviews is a pain when tools only support Mandarin. This article looks at real-world use cases, outlines 4 key factors for choosing a Cantonese transcription tool, compares Tinrec, cSubtitle, Subanana, and SoundType AI, and explains how Tinrec performs in post-meeting summaries, AI Q&A, and team collaboration.