Question · Dictation software
When did the major dictation products launch, and which ones stopped?
Dragon Systems unveiled NaturallySpeaking in 1997, and Microsoft bought Nuance, Dragon's developer, in a deal announced in $19.7 billion. The one clearly ended dictation product in the data is Windows Speech Recognition, deprecated in 2023.
Updated October 25, 20263 min read29 cited facts, each traced to its source
Dictation software started on the desktop and moved into operating systems, then into cloud and open models. Dragon Systems introduced DragonDictate in 1990 1, and unveiled NaturallySpeaking in 1997 2. Few products in this record have stopped. The clearest case is Microsoft's Windows Speech Recognition, which Voice Access replaced.
The full timeline holds 48 data points 3 from 42 sources 4. Each row is a dated event from a maker's own page or a news report of record.
Dragon set the template in the 1990s
DragonDictate came first. Dragon Systems said in 1990 1 that it was the first commercially viable typewriter that works by human speech. That "first" is the company's own claim, and the source does not test it against other products of the time.
NaturallySpeaking followed in 1997 2, described as the first general-purpose, large-vocabulary continuous speech recognition product. Other accounts date retail availability to June of that year, so treat April as the unveiling.
Operating systems made dictation a built-in feature
Microsoft's Windows Vista guide, updated in 2007 5, says speech recognition is part of the operating system. The date is the guide's update month, so the feature is that old or older.
Apple and Google followed. System-wide Mac Dictation arrived with OS X Mountain Lion in 2012 6. Android voice typing was announced for Android 4.0 in 2011 7. Google Docs voice typing arrived in 2015 8. Windows 10 got a dictation hotkey in an Insider build in 2017 9, a preview before any public release.
| Product | Maker | Date | Kind | Year (year) | Source |
|---|---|---|---|---|---|
| Otter.ai Live Assist | Otter.ai | 2026-07-21 | Transcription | 2,026 | Otter.ai, Inc., 2026 10 |
| SpeechAnalyzer and SpeechTranscriber | Apple | 2025-06 | Speech-to-text API | 2,025 | Apple, 2025 11 |
| Nova-3 | Deepgram | 2025-02-12 | Speech model | 2,025 | Deepgram, 2025 12 |
| Google Speech-to-Text Chirp 3 | 2025-04-11 | Speech-to-text API | 2,025 | Cloud, 2025 13 | |
| Voxtral | Mistral AI | 2025-07-15 | Speech model | 2,025 | Mistral AI, 2025 14 |
| Google Speech-to-Text chirp_telephony | 2024-01-08 | Speech-to-text API | 2,024 | Cloud, 2025 15 | |
| Home Assistant Voice Preview Edition | Home Assistant | 2024-12 | 2,024 | project, 2026 16 | |
| Voice Access | Microsoft | 2022-09-20 | Built into an operating system | 2,022 | Windows Experience Blog (Microsoft), 2022 17 |
Showing 8 of 19 rows. See every row in the dataset.
The timeline shows two takeovers and one retirement
Microsoft announced it would acquire Nuance in $19.7 billion 18, and said it had completed the deal on March 4, 2022 19. Microsoft's announcements do not mention Dragon. The link is that Nuance developed it.
The product that stopped is Windows Speech Recognition. Microsoft's list of deprecated features dates its deprecation, and its replacement by Voice Access, to 2023 20. Microsoft's support page then says Voice Access replaced it in the Windows 11 releases of 2024 21. Speech Recognition remains available in older versions of Windows.
| Product | Maker | Date | Kind | Year (year) | Source |
|---|---|---|---|---|---|
| OpenAI transcription models | OpenAI | 2027-02-26 | Transcription | 2,027 | OpenAI API documentation, 2026 22 |
| Google Assistant (Android) | 2026-08-05 | Voice assistant | 2,026 | Droid Life, 2026 23 | |
| Ito | heyito | 2026-01-13 | Dictation app | 2,026 | GitHub, heyito/ito repository (the maker's own repository), 2026 24 |
| Windows Speech Recognition | Microsoft | 2024-09 | 2,024 | Microsoft, 2026 21 | |
| Windows Speech Recognition | Microsoft | 2023-12 | 2,023 | Microsoft, 2026 20 |
| Product | Maker | Date | Kind | Year (year) | Source |
|---|---|---|---|---|---|
| Nuance Communications | Microsoft | 2022-03-04 | Other | 2,022 | Microsoft, 2022 19 |
| Nuance speech recognition | Nuance Communications | 2021-04-12 | 2,021 | Microsoft, 2021 18 | |
| MacSpeech Dictate | Nuance | 2010-02-16 | Dictation app | 2,010 | Speech Technology Magazine, 2010 25 |
| SpinVox | Nuance | 2009-12-30 | Transcription | 2,009 | Rao, 2009 26 |
Recent activity is in models and APIs, not apps
In recent years, most dated events are model and API releases. OpenAI released Whisper as open-source software in September 21, 2022 27. Google's Chirp 3 reached general availability on October 13, 2025 28. NVIDIA's Parakeet v3 and Mistral's Voxtral also appear in 2025 29. The chart shows how events are spread across years.
What the dates do and do not measure
The data is a record of dated announcements, not of market size or quality.
- Announcement is not availability. Several rows are previews, betas or announcements. Siri was called a beta at launch, the Windows hotkey was an Insider preview, and Google Assistant was shown before it reached users.
- Some dates are months only. Dragon, Windows Speech Recognition and Home Assistant's voice hardware carry a month, not a day.
- Some dates are calculated or indirect. Amazon Alexa's month is worked out from a press release that says "seven months ago."
- "First" is usually a vendor claim. Priority statements are not independently tested.
- Coverage is uneven. A product with a well-kept changelog yields many rows. One without yields few. Fewer rows in a year does not mean less happened.
- Absence is not proof. A product missing from the discontinued list may still have ended. The data only shows endings that a maker or news report of record dated.
What this means for choosing how to dictate
Check whether your tool is still developed. If you rely on Windows Speech Recognition, Microsoft deprecated it and points to Voice Access, so plan a move. If you use Dragon, the owner has changed, so check the vendor's current support for your platform before you buy.
Built-in tools from Apple, Google and Microsoft are free and usually enough for short text. Apps built on newer models add things the old desktop products lacked, such as many languages and on-device processing. Yap, which we make, is one on-device option for Mac and Windows. See our comparison of products at /market before you decide.
Keep reading
Evidence
Every number on this page, with where it comes from. Open a fact to see the exact words in its source.
- 1.
- 2.
- 3.
- 4.The timeline of dictation and speech recognition products dataset draws on 42 distinct sources.
Calculated from 48 other facts
- 5.
- 6.
- 7.
- 8.
- 9.
- 10.
- 11.
- 12.
- 13.
- 14.
- 15.
- 16.
- 17.
- 18.
- 19.
- 20.
- 21.
- 22.
- 23.
- 24.
- 25.
- 26.
- 27.
- 28.
- 29.
The data behind these numbers
Go deeper
Related questions
- Is AI dictation better than Dragon NaturallySpeaking?
- Is there a free alternative to Dragon NaturallySpeaking?
- Which Dragon alternatives also control the computer by voice?
- Can you run voice recognition software on a Raspberry Pi?
- Can you upgrade an old Dragon NaturallySpeaking to version 15 or 16?
- Does Apple have a voice to text app?
- Does Dragon NaturallySpeaking need an internet connection?
- Does Dragon NaturallySpeaking use AI?