TypeLess Review (2026) — 92% Accuracy, Cloud-Only

TypeLess Verdict

6.4

out of 10

Accuracy

Cloud

6 Accuracy — Cloud 6 / 10

Speed

Cloud

8 Speed — Cloud 8 / 10

UX

8.4 UX 8.4 / 10

Features

5.6 Features 5.6 / 10

Privacy

4 Privacy 4 / 10

How we score →

Smooth AI-cleanup and translate modes, but aggressive reformatting inflates error counts and privacy is weak

TypeLess version 1.8.0 scores 6.4/10 overall in Voice-list independent testing (tested 2026-06-24). TypeLess Cloud achieves 7.9% aggregate WER across 6 recordings.

Works well for

Excellent filler removal and automatic structuring — output is often ready to paste
Built-in Translate and Ask Anything modes on dedicated hotkeys
Cross-platform: Windows and macOS
Generous 8,000 words/week free tier

Watch out for

LLM post-processing cannot be disabled — adds words and structure not spoken
Audio always uploaded; telemetry every 30 s; no local / privacy mode
Occasional hallucinations: Chinese characters, collapsed output, invented text
$30/mo flat rate — no annual discount that beats Wispr Flow or Aqua Voice

Best for

Users who want polished, structured output without editing — meeting notes, emails, Slack messages

Not for

Users who need verbatim capture or code dictation
Anyone requiring offline / local privacy mode
Casual dictation where LLM reformatting adds unwanted structure

TypeLess Accuracy & Speed

		Model	Accuracy	Speed
English	Cloud	TypeLess Cloud Default Unknown cloud STT provider with mandatory LLM post-processing. Removes fillers, adds punctuation and structure, translates on demand. Cannot be disabled or configured. Single cloud model — no local option available	92.1% Word accuracy The share of words the model got right (100% − word error rate). 100% = every word correct. 7.9% WER Word Error Rate What % of words the model got wrong. 0% = every word correct. 6.5% CER Character Error Rate Same as WER but measured letter-by-letter. Usually lower than WER. 35% PER Punctuation Error Rate How accurately the model placed commas, periods, and other punctuation. 6 / 10	~3s 2–5s range Post-stop latency Seconds from pressing Stop to the final text appearing in your active app. Average across all test recordings. 8 / 10
No models match — turn a filter back on.

Model

Accuracy

Speed

English

Cloud

TypeLess Cloud Default

Unknown cloud STT provider with mandatory LLM post-processing. Removes fillers, adds punctuation and structure, translates on demand. Cannot be disabled or configured.

Single cloud model — no local option available

92.1%

Word accuracy The share of words the model got right (100% − word error rate). 100% = every word correct.

7.9% WER Word Error Rate What % of words the model got wrong. 0% = every word correct. 6.5% CER Character Error Rate Same as WER but measured letter-by-letter. Usually lower than WER. 35% PER Punctuation Error Rate How accurately the model placed commas, periods, and other punctuation.

6 / 10

~3s

2–5s range

Post-stop latency Seconds from pressing Stop to the final text appearing in your active app. Average across all test recordings.

8 / 10

TypeLess for Coding & IT

TypeLess Cloud Cloud

Coding 98.2% 4 err / 219w

Conference 96.7% 7 err / 214w

Coding

All tech terms recognized correctly
Accurate structured list output

"Here is a walkthrough" adds 4 words not spoken
Section headers (Repository, Kubernetes) are LLM inventions
Numbers correct, some bracket formatting differs

Conference

Accurate on accented speaker
All technical terms correct (Parakeet-TDT, whisper.cpp, Georgi Gerganov)
Correct numbers (5-10×, 2022, 2025, 2026)

Added "Parakeet-TDT" hyphen vs spoken "Parakeet TDT"

TypeLess for Everyday & Long-form

TypeLess Cloud Cloud

Casual 98.4% 3 err / 185w

Long-form 89.4% 57 err / 540w

Casual

Zero filler words in output
Clean numbered list structure
All key facts preserved

Minor wording: "3:00 in the morning" vs "3:00 AM"

Long-form

All core statistics correct ($41/$68, 66%, 12k→41k sessions, $9 CPA, 31%)

LLM adds section headers "Paid Acquisition", "Organic Search" not spoken
"For context" and "Comparing a $9 CPA to $68 is a massive gap" are LLM inventions
"Expanding the content team", "Using paid as an accelerant" headers added

alternative wrong extra missing

i want to walk through what we learned this quarter about organic search versus paid acquisition ,because i think the numbers genuinely changed my mind ,and i want the whole team aligned before we set the budget for next year .so ,quick context .for the last three years we have been spending roughly 70% of our marketing budget on paid channels google ads ,a bit of meta ,some linkedin for the enterprise segment .the remaining 30 percent went into content and seo .and the assumption ,honestly ,was that paid is the reliable engine and seo is a slow ,nice-to-have thing on the side .turns out that framing was wrong .let me give you the actual numbers .on paid our blended cost per acquisition climbed from $41 in january to $68 by september .that is a 66% increase in nine months ,and nothing about our targeting changed .the auction just got more expensive ,more competitors bidding on the same keywords ,plus the platform raising minimum bids .meanwhile our organic traffic went from about 12,000 sessions a month to 41,000 . and the cost per acquisition on that channel ,if you amortize the content investment ,was around $9 .$9 versus $68 .that is not a small gap .the honest counterargument is timing : paid converts today you spend one thousand dollars on tuesday you get leads on tuesday .seo is a delayed engine .the articles we published in february did not really start ranking until may or june so there is a real cash-flow difference .if you are a startup that needs pipeline this month ,you cannot just turn off paid and wait two quarters for organic to compound i get that .but here is the thing that surprised me .when we looked at lead quality ,not just volume ,the organic leads had a 31 percent higher trial-to-paid conversion rate .the theory is that someone who finds you by searching for a specific problem is further along in intent than someone who clicks an ad in their feed .they are actively looking so not only is organic cheaper per lead the leads are actually better .so what are we doing differently next year ?three things first ,we are flipping the ratio ,moving to roughly fifty-fifty between paid and organic over the next two quarters ,not all at once ,because we still need the near-term pipeline .second ,we are doubling the content team from two writers to four ,and we are focusing will focus specifically on bottom-of-funnel comparison content ,because that is where the intent and the conversion rates are highest .and third ,we are going to treat paid as an accelerant for content that is already ranking ,instead of a standalone channel .so when an article hits page one organically ,we put paid behind it to compress the timeline .the goal by the end of next year is to get our blended cost per acquisition back under $30 and to have organic driving more than half of all qualified pipeline .right now it is at about 22 percent .that is a big gap to close ,but the trajectory over the last six months tells me it is achievable .anyway that is the short version .we can dig into the channel-level breakdown in a separate session .

TypeLess for Numbers & Structured Data Low accuracy — best result 76.9%

TypeLess Cloud Cloud

Numbers/ITN 76.9% 9 err / 39w

Numbers/ITN

All numbers correct (date, budget, phone, order ID)

Adds "Here are the contact and order details:" header (5 extra words)
Adds numbered list format "1." "2." around phone and order

TypeLess: Noise Resistance

TypeLess Cloud Cloud

Noisy Cafe 93.0% 13 err / 187w

Noisy Cafe

Handles café-noise audio well
Same quality as clean version

"sign up" vs "sign-up" spacing
Minor structural difference in proposal sentence

Tested on Windows 11 26H2 · AMD Ryzen AI 9 HX 370 · 32 GB RAM · NVIDIA RTX 5070 Laptop 8 GB

TypeLess UX & Integration

Getting started & flow

Onboarding flow

Long onboarding; Google OAuth broken during review; translate hotkey shown but non-functional.

3 / 5

Hotkey customization

Custom hotkeys per mode; no PTT, no mouse-button or double-tap trigger.

4 / 5

Error messages

No errors encountered during testing.

5 / 5

Recording experience

Recording overlay UX

Pill expands to full window in Ask Anything mode; clear states for all recording phases.

5 / 5

Stop / cancel UX

Stop and cancel work reliably; no errors encountered.

5 / 5

Text insertion reliability

Inserts into any focused field; shows result overlay when no field is focused.

5 / 5

Auto-insert vs clipboard

No option to restore previous clipboard content after insertion.

4 / 5

Managing your work

Recording history

History tab exists but no search. Filter between Dictation and Ask Anything only. Can save audio as OGG.

3 / 5

Mode / model switching

3 modes (Dictate / Translate / Ask Anything) each on its own hotkey combination.

5 / 5

Idle resource use

~100 MB RAM · 0.5% CPU at rest (cloud).

3 / 5

TypeLess Features

Text processing

AI post-processing

Always-on cloud LLM — filler removal, punctuation, structure. Cannot be disabled.

Custom vocabulary / dictionary

Personal dictionary available (free tier limit not tested).

Text snippets / expansion

Output & extras

Translation mode

Dedicated Translate mode on a separate hotkey — supports 100+ languages.

Ask / Q&A mode

Ask Anything mode: ask questions about selected text or general topics, separate hotkey.

Music auto-mute

File transcription

Export (txt / srt / json)

Voice commands

Local recognition

Offline / local inference

Multiple model options

TypeLess Privacy

TypeLess streams audio to remote servers on every recording. Beyond audio: Telemetry packet sent every 30 s during idle. Claims zero data retention and no training use (stated during onboarding, no toggle available).

Audio uploaded on every recording

Audio sent only after you press Stop

Nothing is uploaded until you confirm by pressing Stop. Cancel before then and the audio never leaves.

Account required

You must create an account (email) to use the app at all — your dictation is tied to an identity.

Sends more than audio

Telemetry packet sent every 30 s during idle. Claims zero data retention and no training use (stated during onboarding, no toggle available).

Opt out of training on your data

Your recordings are not used to train models.

Disable analytics & tracking

Analytics and tracking cannot be fully disabled (e.g. Google Analytics, ad attribution).

Turn off history storage

You can set the app to never store your transcription history.

From the privacy policy not scored

"Zero cloud data retention" — claimed during onboarding, not independently verifiable.
"Never trained on your data" — policy claim, no toggle or audit log.
Telemetry background packets sent every 30 s regardless of recording state.

Pricing

Free $0 Free — 8,000 words per week

8,000 words per week — renews every 7 days
Standard accuracy (same cloud model)
Dictate, Translate, and Ask Anything modes
Personal dictionary, 100+ languages

Subscription $30/mo Pro · $12/mo billed yearly ($144/yr) · $30/mo monthly — no free trial

Unlimited words, enhanced accuracy
Priority access during high demand
Team member management
Early access to new features

Lifetime Not offered

No lifetime / one-time option — subscription only

TypeLess — Pricing — Free (8k words/week) vs Pro ($12/mo yearly, $30/mo monthly) (as of 2026-07-09)

TypeLess on the free tier

How far TypeLess gets you without paying — the basis for its Best free option ranking.

Free limit: 8,000 words per week — renews every 7 days
Account required: Yes — sign-up needed

What you get for free

8,000 words per week — renews every 7 days
Standard accuracy (same cloud model)
Dictate, Translate, and Ask Anything modes
Personal dictionary, 100+ languages

How we judge free tiers →

Methodology

Accuracy scores use WER (Word Error Rate) computed against multi-reference ground truth with {a|b} alternates for valid transcription variants (e.g. 48% and forty-eight percent are both accepted). Audio delivered via virtual cable from ElevenLabs TTS. Single test session on 2026-06-24.

Read the full methodology →

Limitations of this test

TTS source, not human voice — real-world WER will be higher
Single session, no variance measurement across multiple runs
Punctuation (PER) not shown in this table — see raw data
Numbers WER may be overstated for apps that apply ITN (converting spoken to digit form)