Transcription Rate Per Minute
VexaScribe transcription costs between $0.0033 and $0.01 per minute depending on your plan. See detailed rate breakdowns, cost examples, and how pricing compares to alternatives.
Supported formats:
VexaScribe's transcription rate ranges from $0.0033 to $0.01 per minute, depending on your subscription plan. One minute of audio equals one transcription minute. A 1-hour recording costs $0.20 to $0.60
Pricing verified January 2026
The Honest Per-Minute Rate Table (Updated June 2026)
Every major transcription tool, with the actual rate you'd pay and what's included at that price. Speaker labels and language coverage are noted because they often add hidden cost on per-minute APIs.
| Tool | Base rate / min | Speaker labels | Languages | Best for |
|---|---|---|---|---|
| VexaScribe (flat-rate effective) | $0.0033–$0.01 $2–$20/mo plans | Included (all plans) | 99 | Web upload, 200–6,000 min/mo, includes everything |
| OpenAI Whisper API | $0.006 | Not native — needs WhisperX | 99 | Cheap batch API, no diarization needed |
| Deepgram Nova-3 | $0.0036 batch $0.0058 streaming | Included | 30+ | Cheapest at scale (10k+ min/mo), developer pipelines |
| AssemblyAI Best (Universal-1) | $0.012 base +$0.005/min for diarization = $0.017 all-in | Extra ($0.005/min) | 99 | Strong feature set (entity detection, sentiment, summaries) |
| Rev AI | $0.02 (English) | Included | 36 | Accented English, broadcast-grade accuracy |
| Speechmatics | ~$0.005 $0.30/hr enterprise | Included | 50+ | European market, multilingual, enterprise SLA |
| Otter / Notta (flat-rate) | ~$0.08–$0.30 effective $8.33–$30/mo plans | Included | 3–30 (varies by tier) | Meeting recording with calendar integration |
| Whisper local install | $0 forever | Via WhisperX (free, manual setup) | 99 | Private/sensitive content, unlimited volume |
| Rev / GoTranscript (human) | $0.96–$1.99 | Included | English mostly | Legal-grade 100% accuracy (court, medical) |
Prices verified June 2026 from each vendor's public pricing page. We're biased (we built VexaScribe), but the comparison data is honest — Deepgram Nova-3 IS the cheapest per-minute API for pure pay-as-you-go.
Flat-Rate vs Per-Minute — When Does Each Win?
Per-minute pricing looks cheaper because the headline rate is small ($0.006). But at volume, flat-rate monthly plans usually beat it because they include speaker labels, exports, and a web UI. Here's the honest crossover by volume:
| Monthly volume | Cheapest option | Why |
|---|---|---|
| Under 200 min/mo | Pay-as-you-go APIs (Whisper API, Deepgram) | Whisper API at $0.006/min = $1.20 for 200 min. Cheaper than any flat-rate plan's minimum. |
| 200–2,000 min/mo | VexaScribe Starter ($2) or Basic ($5) | Flat plans include diarization + web UI. APIs need separate diarization billing. |
| 2,000–6,000 min/mo | VexaScribe Pro ($10) or Studio ($20) | Studio at $20 = $0.0033/min effective — competitive with Deepgram's base rate, but with included features. |
| 6,000–20,000 min/mo | Deepgram or AssemblyAI API direct | At sustained high volume, per-minute APIs beat flat-rate. Requires developer integration. |
| 20,000+ min/mo | Whisper local install OR enterprise-tier APIs | Whisper local = $0 if you have hardware. Enterprise APIs offer volume discounts. |
The hidden truth: headline per-minute rates often exclude what most users actually need. AssemblyAI's $0.012/min becomes $0.017/min once you add diarization. Whisper API's $0.006/min becomes $0.012+/min once you add a diarization step (WhisperX or similar). Always calculate the all-in rate, not the headline.
What's Included vs Charged Extra (The Hidden Cost Truth)
Per-minute pricing pages advertise the cheapest possible base rate. The actual bill is often 50–100% higher once you add features that most users need. Here's what to watch for:
🎙️ Speaker diarization (who said what)
AssemblyAI charges +$0.005/min on top of base. OpenAI Whisper API doesn't include native diarization — you need to run WhisperX or a separate model. Deepgram, Rev AI, Speechmatics, and VexaScribe include it in the base rate. If you transcribe interviews/meetings, this matters.
🌍 Multi-language support
Whisper API: 99 languages included. Speechmatics: 50+ included. Deepgram: tier-dependent (cheaper tiers have fewer languages). Rev AI: English-strong, 36 total. AssemblyAI: 99 included. If you work with multilingual content, read the language list before signing up.
📦 File size limits
Most APIs: 5 GB per file. Some web tools (Otter free, free converter sites): 25 MB cap — only 5–10 minutes of audio. VexaScribe: 5 GB. Whisper local: unlimited. For lectures, conferences, long podcasts: check the cap first.
⏱️ Real-time vs batch
Live/streaming transcription is more expensive than batch (after-the-fact). Deepgram batch is $0.0036/min, streaming is $0.0058/min — 60% premium. If you don't need live, use batch and save money.
💾 Storage / retention
Most vendors retain transcripts 30–90 days, then auto-delete unless you pay for extended retention. VexaScribe keeps transcripts as long as you want but lets you delete any file at any time. Some enterprise tiers charge for long-term retention separately.
🎯 Model tier upgrades
AssemblyAI has Nano ($0.002/min, lower accuracy) and Best ($0.012/min, Universal-1). Deepgram has Base and Nova-3. The cheapest tier usually isn't the one you want — read accuracy benchmarks before picking.
Transcription Rates: Market Overview
How VexaScribe compares to other transcription options
| Type | Rate/Minute | Rate/Hour | Notes |
|---|---|---|---|
| Human Transcription | $1.00–$3.00 | $60–$180 | Professional services, high accuracy |
| AI (Pay-as-you-go) | $0.10–$0.30 | $6–$18 | Most AI tools without subscription |
| AI (Subscription) | $0.0033–$0.01 | $0.20–$0.60 | VexaScribe and similar services |
Rates are approximate and may vary by provider. VexaScribe rates verified January 2026.
What Does "Per Minute" Mean?
Transcription pricing is based on the duration of your audio file, not processing time. One minute of audio equals one transcription minute.
Important things to know about minute counting:
- A 47-minute recording uses 47 minutes
- A 1-hour video uses 60 minutes
- Silence in recordings still counts
- Multiple speakers don't multiply the cost
VexaScribe Rates by Plan
Higher plans offer lower per-minute rates
Starter
$0.010
/minute
$2/month
200 min included
Basic
$0.0050
/minute
$5/month
1,000 min included
Pro
$0.0040
/minute
$10/month
2,500 min included
Studio
$0.0033
/minute
$20/month
6,000 min included
Cost Examples by Audio Length
See what common transcription jobs cost on each plan
| Audio Length | Starter$0.010/min | Basic$0.0050/min | Pro$0.0040/min | Studio$0.0033/min |
|---|---|---|---|---|
| 30 minutes | $0.30 | $0.15 | $0.12 | $0.10 |
| 1 hour | $0.60 | $0.30 | $0.24 | $0.20 |
| 10 hours | $6.00 | $3.00 | $2.40 | $2.00 |
| 100 hours | $60.00 | $30.00 | $24.00 | $20.00 |
Costs assume you have sufficient minutes in your plan. If you exceed plan minutes, upgrade to continue transcribing.
How VexaScribe Pricing Works
Simple subscription with monthly minute allocations
Choose a Plan
Select a monthly plan with a minute allocation that fits your needs. Start with 30 free minutes.
Upload & Transcribe
Upload audio or video files. Each file deducts its duration from your monthly minutes.
Minutes Reset Monthly
Your minute balance refreshes each billing cycle. Upgrade or downgrade plans anytime.
Billing Details
Important details about how minutes are counted
Rounding
Durations are rounded up to the next minute. A 4:30 file uses 5 minutes.
Silence
Silence in your recordings counts toward duration. We transcribe the full file.
No Rollover
Unused minutes reset each month. They don't carry over to the next billing cycle.
File Limits
Maximum file size is 5GB. Maximum duration is 4 hours per file.
Running Out
If you run out of minutes, existing transcripts remain accessible. Upgrade to transcribe more.
Plan Changes
Upgrade or downgrade anytime. New minutes apply immediately.
Factors That Affect Accuracy
Your rate gets you consistent AI transcription—quality depends on your audio
Audio Quality
Clear recordings transcribe more accurately. 128kbps or higher recommended for MP3s.
Speaker Clarity
Clear speech without mumbling or very fast talking produces better results.
Background Noise
Minimize background noise, music, or multiple overlapping speakers.
Technical Terms
Specialized terminology (medical, legal) may need manual correction.
Ready to Start Transcribing?
Try VexaScribe free with 30 minutes included. No credit card required.
Transcription Pricing FAQ
What is the cheapest transcription per minute in 2026?
OpenAI Whisper installed locally is $0 with unlimited usage (requires Python setup). Among paid cloud APIs, Deepgram Nova-3 at $0.0036/min batch is the cheapest. Whisper API via OpenAI is $0.006/min. For flat-rate monthly plans, VexaScribe Studio at $20/month for 6,000 minutes works out to $0.0033/min — competitive with the cheapest APIs once you use most of your quota. Cheapest base rate ≠ cheapest in practice: speaker labels often cost extra on per-minute APIs (AssemblyAI charges +$0.005/min for diarization), so a $0.012/min tool with labels included may beat a $0.006/min tool with labels billed separately.
Is flat-rate or per-minute pricing cheaper for me?
Depends on monthly volume. Under 200 min/month: pay-as-you-go APIs win (Whisper API at $0.006/min = $1.20 for 200 min). 200-2,000 min/month: VexaScribe Starter ($2/mo for 200 min) or Basic ($5/mo for 1,000 min) typically wins because they include speakers and exports. 2,000-6,000 min/month: VexaScribe Pro/Studio dominates at $0.003-$0.005/min effective rate. Above 10,000 min/month: direct APIs (Deepgram, Whisper) win at scale because you skip the web-tool overhead. Use the comparison table on this page to find your specific crossover point.
Why is Rev AI $0.02/min when Whisper API is $0.006?
Different models and SLAs. Whisper API is OpenAI's own batch transcription using their Whisper model — cheapest because it's volume-pay-as-you-go with no extras. Rev AI uses their proprietary speech model trained on millions of hours of accented English (which is genuinely better for difficult audio) plus better diarization out of the box and an enterprise-grade SLA. AssemblyAI ($0.012/min Universal-1) and Deepgram ($0.0036/min Nova-3) sit between these poles. For clean English, all three deliver 92-97% word accuracy — the price gap reflects extras (diarization, language coverage, support tier), not core ASR quality.
Are speaker labels included in the base per-minute rate?
Varies by vendor. **Included by default:** VexaScribe (all plans), Rev AI, Deepgram. **Charged separately:** AssemblyAI charges +$0.005/min for speaker diarization on top of the $0.012/min base — so the real all-in rate is $0.017/min. OpenAI Whisper API doesn't include diarization natively at all (you'd need to run WhisperX or a separate diarization step). When comparing per-minute rates, always check whether your use case (interviews, meetings, panels) needs speakers and whether they're billed separately.
How does VexaScribe's $20/month compare to per-minute pricing?
VexaScribe Studio at $20/month includes 6,000 minutes — effective rate of $0.0033/min if you use the full quota. At 3,000 minutes/month usage, the effective rate is $0.0067/min (still cheaper than Whisper API). At 1,000 minutes/month, it's $0.02/min (same as Rev AI but with included extras). Flat-rate wins when you can predict volume; per-minute wins when usage is spiky. VexaScribe also includes speaker labels, 99-language support, SRT/VTT/DOCX export, and meeting bot integration — features that cost extra or aren't available on cheaper APIs.
Why is human transcription so much more expensive ($1-$2/min)?
Human transcription requires actual humans listening to and typing the audio. Rev Human at $1.00-$1.99/min, GoTranscript at $0.96/min, TranscribeMe similar — these reflect labor costs. The math: a 1-hour audio file takes a skilled transcriptionist roughly 4 hours to produce a clean verbatim transcript. Even at minimum wage, that's $30+ per audio hour, which is why $60-$120 per audio hour is industry standard. AI is 50-400× cheaper for the same length. Use human transcription only when 100% verbatim accuracy is legally required (court evidence, depositions, medical records for litigation) — and even then, often the workflow is AI draft + human review at $30-$80/hr, not pure human from scratch.
Are there hidden costs I should watch for?
Yes. Common hidden costs across the industry: (1) Speaker diarization charged separately (AssemblyAI +$0.005/min). (2) Custom vocabulary or domain models (some APIs charge extra). (3) Storage beyond 30-90 days (some vendors auto-delete; some bill for long-term retention). (4) Premium support / SLA tier upgrades. (5) Webhook / API call surcharges at very high volume. (6) Real-time/streaming costs higher than batch (Deepgram streaming is $0.0058/min vs $0.0036/min batch). (7) Premium models tier (AssemblyAI "Best" tier costs more than "Nano"). Always read the pricing page's footnotes, not just the headline rate.
What's the cheapest way to transcribe 100 hours per month?
100 hours = 6,000 minutes. The honest comparison: Whisper installed locally on your own machine = $0 forever (requires Python + decent CPU/GPU). Deepgram Nova-3 batch API ≈ $22 (6,000 × $0.0036). VexaScribe Studio = $20 flat. OpenAI Whisper API ≈ $36 (6,000 × $0.006). Rev AI ≈ $120 (6,000 × $0.02). AssemblyAI Best + diarization ≈ $102 (6,000 × $0.017). Rev human transcription ≈ $7,200 (6,000 × $1.20 average). For typical use, VexaScribe Studio gives the best balance — cheaper than direct APIs once you factor in speaker labels, 99 languages, and the web interface that doesn't require developer integration.
What's the difference between flat-rate web tools and pay-as-you-go APIs?
Web tools (VexaScribe, Otter, Notta, Descript) charge monthly flat rates and provide a UI for upload, edit, export. Pay-as-you-go APIs (OpenAI Whisper API, Deepgram, AssemblyAI, Rev AI) charge per minute consumed and require developer integration — you write code to send audio and parse responses. Web tools win for: non-developers, one-time backlogs, easy editing/sharing of transcripts. APIs win for: automated pipelines (e.g., transcribe every new podcast episode), very high volume (10k+ min/month), tight integration with custom apps. Most users want a web tool; only developers building integrations need APIs.
Note: Prices shown are for current subscription plans. Rates may change. Check pricing page for latest information.