> ## Content Index > Fetch the complete content index at: https://www.theaidirectory.com.au/llms.txt > Use this file to discover other available public pages before exploring further. # Resemble AI Review Australia 2026: Developer-grade voice cloning that ditched the simple plan - URL: https://www.theaidirectory.com.au/resemble-ai-review-australia-2026/ - Published: 2026-06-02T23:00:00.000Z - Updated: 2026-10-09T06:35:22.000Z - Description: Resemble AI is voice cloning and deepfake detection built for developers and enterprises, not creators. We tested it for Australians — the new pay-per-use pricing, GST, and who should actually use it. - Author: Adam - Tags: Audio, Voice & Podcasting, Resemble AI, #review **13**/25 ★★★☆☆ Good Resemble AI Verified 1 Oct 2026 Reviewed 3 Jun 2026 Value for Money**3**/5 Ease of Use**2**/5 Feature Depth**4**/5 Australian Fit**2**/5 Reliability & Support**2**/5 From **A$392**/mo · ex-GST · billed in USD Your data No Australian region is advertised. **The gate:** Five stars need 21+ and Australian Fit of at least 4/5. [Skip to the verdict ](#verdict) > **⚠ Repositioned — the voice prices in this review are no longer available.** As at 20 September 2026 `resemble.ai/pricing` is titled **"Deepfake Detection & AI Security Pricing"**. Its plan ladder and its whole rate card now cover Detect, Intelligence, Meetings, Identity and Watermarker. **The text-to-speech, voice-agent, speech-to-text and per-voice cloning rates this review scores are not published anywhere on the pricing page.** > > **Voice has not been discontinued — it has been moved.** Resemble now ships its voice models as **Chatterbox**, MIT-licensed open source you can self-host, with a **Chatterbox Pro** enterprise tier sold through sales. The site files all of it under "Voice AI Research". So the capability is intact and the published price list is gone. > > The pricing section below has been rebuilt around what Resemble actually publishes today. **The 13/25 score was re-adjudicated on 21 September 2026 and it holds.** Value for Money was re-checked against the vendor's current pages and stays at 3/5: the old rate card did not vanish into a sales call, it was replaced by an MIT-licensed open-source release that Resemble's own comparison table prices at **“Free forever”**. What the capability gained in price it lost in purchasability, and the accessibility problem is already scored — at Ease of Use 2/5\. **What this review still needs is a re-scope, not a re-score:** we are reviewing a voice tool at a vendor that now leads with security. > > *Sources: Resemble AI,* [*pricing*](https://www.resemble.ai/pricing?ref=theaidirectory.com.au) *·* [*Resemble TTS*](https://www.resemble.ai/products/text-to-speech?ref=theaidirectory.com.au)*. Read 20 September 2026\. Wayback: capture pending (archive.org unavailable 20 Sep 2026).* > **Verdict (TL;DR):** Resemble AI is a serious voice-cloning and deepfake-detection platform — but it is now built for developers and enterprises, not for the person making their own marketing videos. Some time before this review it dropped its consumer subscription tiers and moved to a pay-per-use "Flex" model where you load credits and pay per second of audio, plus a quote-based Enterprise plan. You get rapid and professional voice cloning, 100 languages, a full API, and a genuinely useful deepfake-detection product. But there is no longer a simple "sign up and narrate a video" plan, the billing is in US dollars, and you have to estimate your own usage. If you are a developer putting a cloned voice into an app, an IVR phone line or a chatbot, it earns a look. If you are a non-technical Australian wanting clean voiceovers, this is the wrong tool — start with our [voice generator roundup](https://www.theaidirectory.com.au/best-ai-voice-generators-australia-2026/) instead. > > *Last reviewed: 21 September 2026 by Adam.* --- ## Who this is for (and who it isn't) Most of the tools we review on this site are for a busy Australian who wants AI to do more with less effort. Resemble AI is one of the exceptions that needs a warning label, because it has quietly stopped being a tool you can just sign up for and use. **Resemble AI is for you if** you are a developer or a product team building a voice into something — an app, an IVR phone system, a chatbot, a game, or a brand voice you call through an API. It is also for media and security teams who need to detect AI-generated audio and video at scale. For those people, it does real work. **Resemble AI is not for you if** you want to type a script, pick a voice, and download a voiceover for a marketing video or a YouTube clip. That used to be possible on a monthly subscription. It is not the product anymore. If you are a 35-year-old Melbourne small-business owner making your own content, this will frustrate you — you will hit an API, a credit meter and a usage estimate before you hit a finished audio file. For that job, read our [best AI voice generators for Australia roundup](https://www.theaidirectory.com.au/best-ai-voice-generators-australia-2026/), where ElevenLabs and Murf are the right starting points. With that out of the way: for the builders Resemble is now aimed at, here is what it does and where the Australian caveats bite. ## What Resemble AI is Resemble AI is a voice-AI platform that does three connected things: it generates and clones voices, it edits and processes audio, and it detects AI-generated (deepfake) audio, image and video. The company positions itself as having built its voice models "with provenance and detection in mind" from the start, which is why detection sits alongside generation rather than as an afterthought ([Resemble AI — homepage](https://www.resemble.ai/?ref=theaidirectory.com.au)). The generation side is the headline. **Rapid Clone** needs about 10 seconds of reference audio and returns a usable clone in under a minute; **Professional Clone** wants 10 to 25-plus minutes of audio, trains for roughly 40 minutes, and produces a voice with a fuller emotional range that is far closer to the source ([Resemble AI — voice creation](https://www.resemble.ai/products/voice-creation?ref=theaidirectory.com.au)). It supports text-to-speech across 100-plus languages and accents, with zero-shot multilingual cloning available in 23 languages through its Chatterbox Multilingual model — meaning you can clone a voice once and generate speech in those languages with accent and emotion retained. The detection side is **Resemble Detect**, with models the company markets at high accuracy (its DETECT-3B-Omni model is advertised at 98.1% audio accuracy), plus a **Watermarker** product that embeds an invisible, persistent marker into generated audio, image and video. Resemble names Netflix, Paramount, Deutsche Telekom and the World Bank among its customers ([Resemble AI — homepage](https://www.resemble.ai/?ref=theaidirectory.com.au)). That is an enterprise and media-security crowd, which tells you who the product is really built for. The thing to understand before anything else: this is now an API-first, developer-and-enterprise product. The simple consumer plan is gone. **Since this review was written, Resemble has gone one step further.** Its public pricing now covers detection and security only, and the voice models have been released as open source under the Chatterbox name, with an enterprise tier sold through sales. The voice capability described below is intact. What you can no longer do is look up what it costs. ## What it costs in Australia **This section was rebuilt on 20 September 2026, and the honest headline is that Resemble no longer publishes a voice price.** The pricing page is now titled "Deepfake Detection & AI Security Pricing". Everything on it — the plan ladder and every line of the rate card — covers detection, intelligence, identity and watermarking. Text-to-speech, voice agents, speech-to-text and per-voice cloning fees have all come off the page. Billing is still US dollars only. AUD figures below are our own conversions at **1 USD ≈ 1.40 AUD, verified 20 September 2026**, shown **ex-GST** per our GST rule. A GST-registered Australian business that supplies its ABN is not charged GST on an imported digital service; other buyers should allow about 10% on top. | Plan | USD/mo (billed annually) | AUD/mo (ex-GST, \~1.40) | Seats | Notes | | -------------- | ------------------------ | ----------------------------- | ------ | ---------------------------------------------------------------------------- | | **Flex** | $0 | $0 | 1 | Pay-as-you-go credits that never expire. No subscription fee | | **Team** | $280 (US$350 monthly) | \~A$392 (\~A$490 monthly) | 5 | Lower usage rates, batch and large-file uploads | | **Business** | $800 (US$1,000 monthly) | \~A$1,120 (\~A$1,400 monthly) | 20 | SSO, org-wide calendar integration, security notifications | | **Enterprise** | Custom quote | Custom | Custom | Volume pricing, SLAs, SOC 2 documentation, custom model training, on-premise | **Processing rates**, all per second unless stated: | Service | Flex (USD) | Team / Business (USD) | Flex (AUD ex-GST, \~1.40) | | ------------------ | -------------- | --------------------- | ------------------------- | | Detect — audio | $0.03500 | $0.01500 | \~A$0.049 | | Detect — image | $0.03500/image | $0.01500/image | \~A$0.049/image | | Detect — video | $0.07000 | $0.03000 | \~A$0.098 | | Intelligence | $0.02500 | $0.01500 | \~A$0.035 | | Identity search | $0.00050/call | $0.00050/call | \~A$0.0007/call | | Watermarker encode | $0.00050/call | $0.00050/call | \~A$0.0007/call | | Watermarker decode | $0.00020/call | $0.00020/call | \~A$0.0003/call | *Pricing as at 20 September 2026\. Source:* [*resemble.ai/pricing*](https://www.resemble.ai/pricing?ref=theaidirectory.com.au)*. Rate basis: 1 USD ≈ 1.40 AUD, verified 20 September 2026; figures shown ex-GST per our GST rule. Resemble publishes no AU-specific GST guidance, so confirm the tax line on your first invoice.* **What happened to the voice prices.** In June this review recorded text-to-speech at US$0.0005 a second, voice agents and speech-to-text at US$0.001, Rapid clones at US$2 a voice a month, Professional clones at US$5, voice design at US$2 and team seats at US$20\. **None of those figures is on the site today.** What Resemble publishes instead is **Chatterbox**, its TTS and zero-shot cloning models, released **MIT open source** and free to self-host, alongside a **Chatterbox Pro** enterprise tier priced by sales. Deepfake detection for audio, meanwhile, came down from US$0.04 to US$0.035 a second on Flex, and to US$0.015 on Team and Business. **Value verdict:** We are not going to pretend we can price this for you. If you want detection, the numbers above are published, complete and legible — and at A$0.049 a second for audio on Flex, an hour of material costs about A$177 before you have detected anything at volume. If you want voice, there is no self-serve price at all: you either run Chatterbox yourself under its MIT licence, or you go through sales for Chatterbox Pro. That is a defensible model for a developer. It is a dead end for anyone who wanted to know what a month costs before signing up. *Pricing as of 3 June 2026\. Source:* [*resemble.ai/pricing*](https://www.resemble.ai/pricing?ref=theaidirectory.com.au)*. Resemble publishes no AU-specific GST guidance, so confirm the exact GST line on your first invoice and check the tax invoice carries an ABN reference for your accountant. Per-second rates are the vendor's published figures; the AUD-per-hour conversions are our own rounding for context.* **Value verdict:** The per-second rates look genuinely cheap on paper — plain text-to-speech at around AUD$2.80 an hour before tax is not much. But "cheap per second" is not the same as "predictable", and that is the catch for an Australian buyer. You cannot walk in and know what a month will cost; you have to forecast your own usage in seconds of audio, then add per-voice monthly fees for every clone you keep active and a USD$20 seat for every team member. For a developer with a known workload that is fine — arguably better than a fixed subscription. For anyone who wants to know their bill in advance, it is an unknown until you have run it. And the headline detection rate (AUD$0.068 a second for audio, more for video) adds up fast at any real volume. ## How it works in practice You start by loading credits — the Flex plan lets you begin at zero — and then you work through either the dashboard or, more often, the API and SDKs. To clone a voice you upload reference audio: 10 seconds for a Rapid Clone that is ready in under a minute, or 10 to 25-plus minutes for a Professional Clone that trains for around 40 minutes and comes back markedly more natural ([Resemble AI — voice creation](https://www.resemble.ai/products/voice-creation?ref=theaidirectory.com.au)). Every generated file is watermarked at creation, before it leaves Resemble's infrastructure. From there you generate speech from text, convert one voice to another (speech-to-speech), edit audio, or dub into other languages. The multilingual story is the strong part: clone once, then produce the same voice in 23 languages through Chatterbox Multilingual with the accent and emotion carried across, or reach for text-to-speech across 100-plus languages and accents. The detection workflow is separate and aimed at a different user. You send audio, image or video to Resemble Detect and get back a verdict on whether it is AI-generated, billed per second. There is also a Chrome extension for the deepfake detector. This is the part media, trust-and-safety and security teams care about, and it is the clearest reason an enterprise rather than a creator ends up here. What you will notice quickly is that none of this is "open the app and finish a task in five minutes." The centre of gravity is the API. If you are not comfortable making API calls or wiring an SDK into your own software, you will feel like you have walked into the wrong shop — because, for the consumer use case, you have. ## What it does well - **Fast, flexible voice cloning with a real quality tier.** Rapid Clone in under a minute from 10 seconds of audio is genuinely quick; Professional Clone trades time for fidelity. Having both, rather than one compromise setting, is the right design for a developer. - **Strong multilingual coverage.** 100-plus languages for text-to-speech and 23-language zero-shot cloning that keeps the voice's accent and emotion is more than most rivals offer, and it matters for anyone building for more than an English-only audience. - **Detection built in, not bolted on.** Resemble Detect and the Watermarker sit on the same platform as generation. For a media or security team worried about deepfakes, having creation and detection from one vendor is a coherent story, not a marketing line. - **A pay-per-use model that can be cheap at the right volume.** No fixed subscription and credits that never expire mean a developer with a modest or spiky workload only pays for what they use. - **The voice models are now MIT open source.** Chatterbox and Chatterbox Multilingual can be self-hosted at no licence cost, which is a genuine option for a team with infrastructure and a data-residency problem. - **Serious enterprise plumbing.** SOC 2, SSO/SAML, custom model training and on-premise deployment on the Enterprise plan, with named customers like Netflix and Paramount. This is built to pass a procurement review. ## What it doesn't do well - **There is no simple plan for non-technical users anymore.** This is the big one. Dropping consumer subscriptions in favour of an API-first Flex model means a small-business owner who just wants a voiceover has nowhere obvious to land. If that is you, this tool is not for you. - **You cannot know your bill in advance.** Pay-per-use means estimating usage in seconds of audio. Fine for a developer with a forecast; frustrating for anyone who wants a fixed monthly number. - **There is no published price for the voice product at all.** As at 20 September 2026 the pricing page covers detection and security only. For voice you self-host Chatterbox or you talk to sales — there is no figure to compare against ElevenLabs or Murf. - **USD-only pricing.** No AUD anywhere on the site. You are charged in US dollars at your bank's FX rate, with GST added at an Australian checkout, so the landed cost runs above the headline per-second figure. - **No Australian data residency.** Resemble is a US-based vendor with no advertised AU hosting region. For voice cloning of real Australians — which can be sensitive personal information — and for anything touching the Privacy Act, that offshore-data question needs deliberate thought before you upload reference audio. - **No local Australian support.** No AU office, no AU business hours, no phone line outside an Enterprise agreement. Dedicated support is an Enterprise-tier feature, not something a Flex user gets. ## Australian fit - **Available in Australia:** Yes — no geographic restriction. Self-serve signup and the Flex plan work globally. ([resemble.ai/pricing](https://www.resemble.ai/pricing?ref=theaidirectory.com.au)) - **Pricing in AUD:** No. USD only. The published rates now cover detection and security; there is no published voice price. Expect FX on top, and about 10% GST unless you supply an ABN. - **GST:** Not stated on the pricing page. Expect around 10% added at checkout under Australia's imported-digital-services rules — confirm on your first invoice and check it is a compliant tax invoice carrying an ABN reference for your accountant. - **Data residency:** No Australian region is advertised. US-based vendor; cloning a real person's voice means uploading their audio offshore — material under the Privacy Act's sensitive-information provisions deserves caution. Enterprise offers on-premise deployment on quote, so any region control is a contract-level conversation rather than a published option. - **Compliance:** SOC 2 and SSO/SAML are Enterprise-tier features, with on-premise deployment available on quote. No IRAP, no AU-specific certification advertised. ([resemble.ai/pricing](https://www.resemble.ai/pricing?ref=theaidirectory.com.au)) - **AU-aware output:** Resemble supports 100 languages and regional dialects (with full zero-shot voice cloning in 23 of them), but advertises no Australian-accent-specific control the way some rivals do ([Resemble — Text to Speech](https://www.resemble.ai/products/text-to-speech?ref=theaidirectory.com.au)). - **Local AU support:** None. Dedicated support is Enterprise-only; Flex users rely on docs and standard channels. No AU hours. ## Alternatives | App | Why you might prefer it | | ---------------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | | **ElevenLabs (API)** | The strongest all-round voice quality and the more mature developer API, with AU-accent control and a free tier to start. The default first look for most voice-cloning builds. | | **Murf** | If you are not a developer and want a calm, predictable interface to produce voiceovers without watching a per-second meter, Murf is the easier buy. | | **For non-developers** | If you just want to narrate a video and you do not write code, skip the API-first tools entirely — read our [best AI voice generators for Australia roundup](https://www.theaidirectory.com.au/best-ai-voice-generators-australia-2026/) and start with ElevenLabs or Murf. | ## Bottom line Resemble AI is a capable platform that has made a deliberate choice about who it serves. By dropping consumer subscriptions and going API-first with a pay-per-use Flex model and a quote-based Enterprise tier, it has become a better tool for developers and a worse one for everyone else. The voice cloning is fast and tiered sensibly, the 100-language coverage is real, and the built-in deepfake detection is a genuine differentiator for media and security teams. But this publication writes for busy Australians who are mostly not developers, and for that reader the honest answer is: this is no longer your tool. There is no simple plan, the billing is in US dollars you have to forecast yourself, there is no Australian data residency, and there is no local support. Those are the reasons it scores 13 and sits in the "Good — for the right buyer" band rather than higher. If you are a developer or product team cloning voices via API at scale, or a team that needs deepfake detection alongside generation, try the Flex plan with a small credit load and model your real usage before you commit. If you are anyone else making voiceovers, our voice roundup will serve you better. We would expect to revisit this inside six months — the pricing model is new enough that it may shift again. ## Score | Criterion | Score | Notes | | --------------------- | --------- | ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | | Value for Money | 3/5 | **Re-checked 21 Sep 2026 — holds.** The voice models are now **Chatterbox, MIT-licensed and free to self-host**, and the detection side offers Flex at US$0/mo with non-expiring credits and no card — two genuine free paths, plus a real 20% annual discount. Held at 3, not raised, because there is **no published self-serve paid voice tier at all** (Chatterbox Pro is contact-sales) and no AUD billing, so a buyer cannot compare or purchase | | Ease of Use | 2/5 | API-first with no simple consumer plan; a non-technical Australian cannot easily get to a finished file | | Feature Depth | 4/5 | Rapid and Professional cloning, 100 languages, speech-to-speech, dubbing, plus a real deepfake-detection and watermarking product | | Australian Fit | 2/5 | USD-only pricing, no AU data residency, no AU support, GST unstated; offshore voice-cloning data is a real consideration | | Reliability & Support | 2/5 | SOC 2, named enterprise customers and on-premise options are reassuring, but dedicated support is Enterprise-only and there are no AU hours | | **Total** | **13/25** | ⭐⭐⭐ Good — re-adjudicated 21 September 2026, unchanged | Update history - **1 Oct 2026 — re-verified by browser; negative confirmed, ladder clean.** `resemble.ai/pricing` on an Australian connection: Flex free (pay as you go), Team $280 and Business $800 a month billed annually (the stated savings imply $350 and $1,000 monthly), Enterprise custom. Every plan figure holds; per-use detection rates were not re-read. Bare `$`, no `A$`. `aud_native: false` set. Score unchanged. Capture: `Dossiers/captures/resemble-ai-pricing-2026-10-01.txt`; [Wayback 1 Oct 2026](https://web.archive.org/web/20260930193937/https://www.resemble.ai/pricing). - **21 Sep 2026** — **Value for Money re-checked and held at 3/5\. The total stays 13/25 and no criterion moved.** The 20 September pass flagged the criterion on the grounds that the voice rate card had disappeared. Re-verified against Resemble's own pages today, the opposite is true for the thing this review is about: the TTS and cloning models now ship as **Chatterbox under an MIT licence**, and Resemble's comparison table on [resemble.ai/chatterbox](https://www.resemble.ai/chatterbox/?ref=theaidirectory.com.au) prices it at **"Free forever"** against ElevenLabs at US$0.15/1K characters, OpenAI TTS at US$15/1M and Azure TTS at US$24/1M, with on-premise deployment and zero-shot cloning from five seconds of audio. **We scored a metered rate card; what replaced it is free.** Against that, the self-serve voice subscription is gone — there is no paid voice tier a reader can buy, only a repository to clone or a sales team to ring. Free-but-unpurchasable and metered-but-purchasable land in the same place, and the accessibility loss is already priced at Ease of Use 2/5, so charging it twice would be wrong. **The open question is scope, not score:** this is a voice review at a vendor that now sells deepfake detection. Filed for a re-scope. Full reasoning: `Audits/contested-scores-2026-09-21.md`. [Source](https://www.resemble.ai/chatterbox/?ref=theaidirectory.com.au) · [Pricing](https://www.resemble.ai/pricing?ref=theaidirectory.com.au) · Wayback: capture pending (archive.org unavailable 21 Sep 2026) - **20 Sep 2026** — **Resemble's pricing page now sells deepfake detection, and the voice rates this review is built on are no longer published.** The page is titled "Deepfake Detection & AI Security Pricing". **Old, now unpublished: text-to-speech US$0.0005/sec, voice agents US$0.001/sec, speech-to-text US$0.001/sec, Rapid clone US$2/voice/mo, Pro clone US$5/voice/mo, voice design US$2/voice/mo, team seat US$20/mo.** **New ladder: Flex US$0 pay-as-you-go (1 seat) · Team US$280/mo billed annually, US$350 monthly (5 seats) · Business US$800/mo billed annually, US$1,000 monthly (20 seats) · Enterprise custom.** **Deepfake detection for audio is now US$0.035/sec on Flex and US$0.015/sec on Team and Business, was US$0.04/sec.** [Source](https://www.resemble.ai/pricing?ref=theaidirectory.com.au) · Wayback: capture pending (archive.org unavailable 20 Sep 2026) - **20 Sep 2026** — **The voice product moved to open source.** Resemble now publishes its TTS and cloning as **Chatterbox** under an MIT licence, with **Chatterbox Pro** sold as a contact-sales enterprise tier. There is no self-serve voice price to quote. That removes the per-second economics **Value for Money 3/5** rested on, so the criterion and the 13/25 total are **flagged for re-rating**. [Source](https://www.resemble.ai/products/text-to-speech?ref=theaidirectory.com.au) · Wayback: capture pending (archive.org unavailable 20 Sep 2026) - **3 June 2026** — First reviewed. ## Frequently asked questions ### Is Resemble AI available in Australia? Yes. There's no geographic restriction — self-serve signup and the Flex plan work globally. There's no AU office or local support, though. ### Does Resemble AI charge in AUD? No. Pricing is USD only. As at 20 September 2026 the published rates cover deepfake detection and security — Flex US$0 pay-as-you-go, Team US$280/month billed annually, Business US$800 — and there is no published voice price at all. Expect your bank's FX rate on top, plus about 10% GST unless you supply an ABN. ### Is there a free plan or free trial? The pay-per-use Flex plan lets you start at $0 — load credits when you need them, and they never expire. There's no fixed free tier, but you only pay for what you use. Resemble's voice models are also released as Chatterbox under an MIT open-source licence, so you can self-host them at no licence cost. ### Does Resemble AI store data in Australia? No Australian region is advertised. It's a US-based vendor, so cloning a real person's voice means uploading their audio offshore — a real consideration under the Privacy Act's sensitive-information provisions. On-premise deployment is Enterprise-only, on quote. ### Who is Resemble AI best for? Media and security teams needing deepfake detection, which is what Resemble now prices publicly. Also developers willing to self-host its open-source Chatterbox voice models, or to negotiate an enterprise voice contract. Non-technical Australians wanting simple voiceovers should look at ElevenLabs or Murf instead. ## Read next - Category guide: [Best AI Voice Generators in Australia 2026](https://www.theaidirectory.com.au/best-ai-voice-generators-australia-2026/) - [Adobe Podcast review](https://www.theaidirectory.com.au/review-adobe-podcast-australia-2026/) - [Cleanvoice review](https://www.theaidirectory.com.au/review-cleanvoice-australia-2026/) ## Sources 1. Resemble AI — pricing page, titled "Deepfake Detection & AI Security Pricing" (Flex US$0 / 1 seat · Team US$280 annual, US$350 monthly / 5 seats · Business US$800 annual, US$1,000 monthly / 20 seats · Enterprise custom; Detect audio US$0.035 Flex and US$0.015 Team/Business per second; Detect video US$0.07 / US$0.03; Intelligence US$0.025 / US$0.015; Identity search US$0.0005/call; Watermarker encode US$0.0005, decode US$0.0002; no voice, TTS or cloning rate published) — [resemble.ai/pricing](https://www.resemble.ai/pricing?ref=theaidirectory.com.au) — Wayback: capture pending (archive.org unavailable 20 Sep 2026) — checked 20 September 2026 1a. Resemble AI — Resemble TTS product page (Chatterbox and Chatterbox Multilingual MIT open source; Chatterbox Turbo; Chatterbox Pro enterprise tier sold via "Contact us"; filed under "Voice AI Research") — [resemble.ai/products/text-to-speech](https://www.resemble.ai/products/text-to-speech?ref=theaidirectory.com.au) — Wayback: capture pending (archive.org unavailable 20 Sep 2026) — checked 20 September 2026 1b. Resemble AI — pricing page, superseded (Flex model, per-second voice rates, voice-clone and seat add-ons, Enterprise) — [resemble.ai/pricing](https://www.resemble.ai/pricing?ref=theaidirectory.com.au) — checked 3 June 2026 2. Resemble AI — homepage (Resemble Detect, DETECT-3B-Omni, Watermarker, named customers, provenance positioning) — [resemble.ai](https://www.resemble.ai/?ref=theaidirectory.com.au) — checked 3 June 2026 3. Resemble AI — voice creation product (Rapid vs Professional Clone timings, watermarking at creation) — [resemble.ai/products/voice-creation](https://www.resemble.ai/products/voice-creation?ref=theaidirectory.com.au) — checked 3 June 2026 4. Resemble AI — text-to-speech / multilingual product (100 languages and regional dialects; 23-language Chatterbox Multilingual zero-shot cloning) — [resemble.ai/products/text-to-speech](https://www.resemble.ai/products/text-to-speech?ref=theaidirectory.com.au) — checked 3 June 2026 *Conversion note: AUD ≈ 1.40 × USD, rate verified 20 September 2026\. GST assumption: most international SaaS displays ex-GST USD and adds \~10% GST at an Australian checkout under imported-digital-services rules; AUD figures are indicative — confirm at checkout..* --- *The AI Directory Australia is an independent editorial publication. We don't accept payment for positive reviews. Read our Methodology and Affiliate Disclosure pages.* *Reviewed by* [*Adam*](mailto:adam@theaidirectory.com.au) *for* [*The AI Directory Australia*](https://www.theaidirectory.com.au/)*. First published 3 June 2026.* ---