Australian Transcription vs Otter.ai, Sonix & AWS Transcribe
What Australian businesses actually need from a transcription service — and how the options stack up.
The Australian context matters
Most transcription services are built for a global market. That's fine for many use cases. But Australian businesses operating under the Privacy Act 1988 (Cth) have a specific problem: APP 8 — the cross-border disclosure principle.
APP 8 requires that before you disclose personal information to an overseas recipient, you take reasonable steps to ensure that recipient won't breach the Australian Privacy Principles. In practice, this means sending audio recordings overseas to US-hosted services carries real legal exposure — not just theoretical risk.
Australian Transcription processes all audio on AWS infrastructure in Sydney. Your data never leaves Australia, so APP 8 obligations are never triggered. It's the only transcription API built around this constraint from the ground up.
With that framing in mind, here's how the main options compare.
Side-by-side comparison
| Feature | Australian Transcription | Otter.ai | Sonix | AWS Transcribe |
|---|---|---|---|---|
| HQ | Australia | USA | USA | USA |
| Data location | Australia (always) | USA | USA |
Configurable Requires manual region config |
| APP 8 compliance | Obligation never triggered | Triggered by cross-border disclosure | Triggered by cross-border disclosure | Requires manual AU region setup |
| Accuracy | 93.5% | Not disclosed | ~92% | 92.4% |
| Pricing (AUD) |
$0.02/min Speaker diarization included |
Subscription required No pay-per-minute option |
US$10/hour Pay-per-minute |
~$0.042/min Pay-per-minute; diarization included |
| Australian spelling | Default | No | No | Yes |
| Speaker diarization | Included | Available | Available | Available |
| Immediate auto-delete | Default | No | No | Requires coding |
| Free tier | 90 min free, no credit card | Limited free plan (subscription) | 30 min free trial | 60 min/month on AWS Free Tier |
| API access | Yes | Limited (primarily web UI) | Yes | Yes (AWS SDK) |
Competitor pricing is based on publicly listed rates and was last verified in August 2026. Rates, regions and included features change; confirm current terms with each provider before making a purchasing decision.
Pricing shown in AUD based on publicly listed rates. Verify current rates at each provider before making purchasing decisions. Accuracy figures from publicly available benchmarks.
Otter.ai
Otter.ai is primarily a meeting assistant, not a transcription API. It's good at real-time capture for Zoom and Google Meet, but it's subscription-based, stores recordings on US servers, and doesn't offer straightforward pay-per-minute API access. For Australian businesses needing programmatic transcription with data sovereignty, it's not the right fit.
Sonix
Sonix has strong editing tools and a clean web interface, making it popular for journalists and content creators. It does offer an API and pay-per-minute pricing, but at US$10 per audio hour it's considerably more expensive than the alternatives. Data is stored on US servers, so APP 8 obligations apply when processing audio involving Australian individuals.
AWS Transcribe
AWS Transcribe is the closest competitor on data residency — you can configure it to use the ap-southeast-2 (Sydney) region. It also supports Australian English spelling. But "configurable" means it won't be Australian by default: you need to explicitly set the region, and you'll need an AWS account, IAM setup, and SDK integration. It's a reasonable choice for teams already deep in AWS. For everyone else, the overhead is significant compared to a purpose-built API with Australian residency by default.
On pricing, AWS Transcribe comes in around $0.042/min (AUD), which is more expensive than Australian Transcription's flat $0.02/min. Speaker diarization is included in both rates, so that is a straight comparison.
Choosing between them
The four services above are not really competing for the same job, which is why head-to-head accuracy comparisons tend to mislead. Work out which of these describes you first, and the shortlist collapses quickly.
You want something to sit in your meetings and take notes. That is a meeting assistant, not a transcription API. Otter.ai is built for exactly this and does it well. You are buying a product with an interface, not a service you integrate. The trade-off is that your recordings live on US infrastructure and you are on a subscription rather than paying for what you use.
You produce transcripts as a deliverable. Journalists, researchers, and content teams who spend real time correcting and formatting transcripts benefit from a proper editing environment. Sonix is the strongest option here. It is also by far the most expensive per minute, which is defensible if the editing tools save you an hour and irrelevant if you never open them.
You are building transcription into software. Now you want an API, predictable per-minute pricing, and a residency answer you can put in a tender response. This is where AWS Transcribe and Australian Transcription actually compete. AWS wins if you are already running infrastructure there and want one bill and one IAM model. We win if you would rather not stand up an AWS account, configure a region correctly, and pay separately for diarization to get the same outcome.
You need it live, as the audio happens. None of these is the right answer. Real-time streaming is a different product category — look at a streaming-first API instead. We are asynchronous only and say so plainly.
The hidden costs worth counting
Per-minute rates are the visible number, and they are rarely the number that decides the bill. Three things tend to matter more.
Per-feature billing. Speaker diarization is charged as an extra by some API providers, though not by AWS. If you need to know who said what — and for interviews, meetings and focus groups you almost always do — check whether the headline rate is the rate you actually pay. Ours includes it.
Currency exposure. US-listed services bill in USD. Your budget is in AUD. A rate that looked cheap when you signed off on it moves with the exchange rate, and the variance is not small over a year.
Compliance overhead. If your audio contains personal information about Australians and it is processed overseas, someone has to do the APP 8 assessment, update the privacy policy, and answer the residency question in every tender and security review from now on. That is recurring work with a real salary cost attached, and it does not appear on any pricing page.
Compare individually
For a closer look at any single service, including migration notes and side-by-side code:
Frequently asked questions
Which transcription service is best for Australian businesses?
It depends on what you are optimising for. If you need programmatic access and your audio contains personal information about Australians, an Australian-hosted API avoids APP 8 cross-border obligations entirely. If you want a meeting assistant that joins calls and takes notes, Otter.ai does that and we do not. If you need a polished editing interface for producing publication-ready transcripts, Sonix is strong. If you are already deep in AWS, Transcribe in ap-southeast-2 is a reasonable choice. There is no single best answer, only a best fit.
Do any of these services keep data in Australia?
AWS Transcribe can, if you explicitly configure the ap-southeast-2 (Sydney) region — but it is not the default, and misconfiguring it sends your audio elsewhere. Australian Transcription processes everything in Sydney with no configuration required. Otter.ai and Sonix both process and store on US infrastructure.
How much does transcription cost in Australia?
Australian Transcription is $0.02 AUD per minute with speaker diarization included, which is $1.20 per audio hour. AWS Transcribe is roughly $0.042 AUD per minute, also with diarization included. Sonix publishes US$10 per audio hour, about $0.17 USD per minute. Otter.ai is subscription-based rather than per-minute, so the effective rate depends entirely on how much you use it. Human transcription in Australia runs $1 to $2 per audio minute.
Does any of them handle Australian accents and spelling properly?
Australian accents are handled well by modern speech models across all of these services — accent is much less of a differentiator than it was five years ago. Spelling is a different matter: services that default to US English will give you 'organize' and 'color' unless you set the locale. AWS Transcribe supports en-AU explicitly, and our output uses Australian spelling by default.
Can I use a US transcription service and still comply with the Privacy Act?
Yes, but it takes work. APP 8 does not prohibit sending personal information overseas — it makes you accountable for what happens to it. You need to take reasonable steps to ensure the overseas recipient complies with the Australian Privacy Principles, disclose the likely destination countries in your privacy policy, and in most cases you stay liable for their breaches. Using an Australian-hosted service means the obligation is never triggered in the first place.
What should I actually test before choosing?
Run the same five to ten real recordings through each shortlisted service, including your worst audio rather than your best. Compare how many proper nouns and technical terms survive, whether speaker labels stay consistent, and how much editing each transcript needs. Then price it against your real monthly volume in AUD. Published accuracy benchmarks are measured on audio that sounds nothing like yours.
Try it free before committing
Sign up and get 90 minutes of free transcription. No credit card required. Test it on your own recordings before deciding.