# Bland Speech vs Fish Audio voice cloning (2026): Who else gets to use a cloned voice?

On Fish Audio, a professional clone can be released so that other creators pay to use it, and the person who made it earns a share. On Bland Speech, every clone is private to the customer's organization. The two companies built cloning for different owners, and the rest of the differences follow from that.

Fish Audio is a voice AI platform for creators, developers, and enterprises, with a web app, an API, open model weights, and a self-hosted tier. Bland Speech is a text to speech model for production phone calls. A brand cloning its own spokesperson for a call line and a creator cloning a voice to sell are shopping for different products.

## Which is better, Bland Speech or Fish Audio for voice cloning?

For a business cloning its own voice for phone calls, Bland Speech is the better choice over Fish Audio as of November 2026. Clones stay private to the organization, output includes mu-law at 8 kHz for phone lines, and Bland Speech v3 ranked first of 13 AI voices in Design Arena's Audio Realism Bench results of 4 August 2026.

Fish Audio makes more sense for a creator or developer who wants a lower per-byte price, 83 languages on one model, and the option to run open weights on a team's own hardware. A team choosing among more providers can start from the [voice cloning comparison](/compare/text-to-speech/voice-cloning) or the [text to speech comparison](/compare/text-to-speech).

<div class="cmp-usewhen"><div class="cmp-usewhen-own"><p><strong>Use Bland Speech when</strong></p><ul><li>A brand voice should stay private to the organization that cloned it</li><li>A team will create an instant clone through an API from one sample of about 10 seconds</li><li>A professional clone should pass a consent and rights review before it trains</li><li>Cloned speech will travel over a phone line as mu-law at 8 kHz</li><li>A team wants one standard per-character rate with no monthly plan to manage</li></ul></div><div><p><strong>Use Fish Audio when</strong></p><ul><li>A creator will release a professional voice for other creators to pay to use, and earn a share</li><li>A team will build a professional clone in a web app, from 10 to 180 minutes of audio, on a paid plan</li><li>A product needs cloned voices across 83 languages on one model</li><li>A team will weigh a self-hosted tier, with VPC, on-prem, air-gapped, or sovereign cloud deployment, against a $10K monthly minimum.</li><li>Delivery should be steered with natural-language bracket tags such as [whispers sweetly]</li></ul></div></div>

| Feature | Bland Speech | Fish Audio |
|---|---|---|
| Audio Realism Bench, 4 August 2026 | Ranked 1st of 13 AI voices, 1365 | Wasn't in those results |
| Instant clone sample | One clean sample, about 10 seconds | At least 10 seconds per clip, and a minute or two improves fidelity |
| Professional clone | 30 minutes or more, after a consent and rights review | At least 10 minutes of audio, up to 180, built in the web app |
| Who can use a clone | Private to the customer's organization | Released professional voices can earn a share when other creators pay to use them |
| Plan | Pay as you go, no monthly plan | API pay as you go, and professional clones on the Plus or Pro app plans |
| Price for speech | $0.04 per 1,000 characters ($40 per million) | $15 per million UTF-8 bytes on s2.1-pro, $50 on drama-3 |
| First audio, as each company publishes it | 199ms | Fish claims under 70 ms latency, and 100ms to first audio on s2-pro |
| Phone audio | Mu-law at 8 kHz, or PCM from 8 to 48 kHz | Output as mp3, wav, pcm, or opus |

<details><summary>What these rows mean</summary>

- **Audio Realism Bench, 4 August 2026:** Design Arena's blind listening test, where listeners hear two clips of the same line and pick the one that sounds more like a real person. Fish Audio wasn't in those results, and every score in them is for a stock voice, not a clone.
- **Instant clone sample:** An instant clone copies a voice from a short clip. Fish Audio accepts wav, mp3, m4a, and opus clips.
- **Professional clone:** A professional clone is trained on much more audio. Fish Audio says the whole flow takes about 1 to 2 hours, with training of 10 to 30 minutes, and a professional clone needs a paid plan, with 1 slot on Plus and 5 on Pro.
- **Who can use a clone:** Fish Audio says a released professional voice earns a share when other creators pay to use it. On Bland Speech the customer confirms each voice is theirs to use, and clones stay private.
- **Price for speech:** Fish Audio bills per million UTF-8 bytes, which tracks characters for English text and runs higher for scripts that take more bytes per character. Fish converts a million bytes to about 180,000 English words.
- **First audio:** Time to first audio is the wait before sound starts. Each figure is the company's own claim.
- **Phone audio:** Mu-law is the 8-bit format US phone networks carry.

</details>

## How do Bland Speech and Fish Audio clone a voice?

Bland Speech and Fish Audio both clone from a short clip, and both have a longer route for a voice that has to be exact. They differ on where that longer route runs and on who is allowed to use the result.

### Bland Speech overview

Bland Speech comes from Bland, a voice AI provider that runs its own speech model for enterprise phone calls and sells it as a studio and an API on the [Bland Speech](/speech) page. The [text to speech API](/text-to-speech-api) page has the endpoints.

Cloning has two tiers. An instant clone is made from one clean sample of about 10 seconds, and a professional clone is fine-tuned on 30 minutes or more of audio after Bland's consent and rights review. Every clone stays inside the customer's organization, and the customer confirms each voice is theirs to use.

<div class="cmp-proscons cmp-proscons-own"><div class="cmp-proscons-pros"><p><strong>Pros</strong></p><ul><li>An instant clone from about 10 seconds, created through an API</li><li>A professional clone after a consent and rights review</li><li>Clones private to a customer's organization</li><li>Mu-law and PCM output from 8 to 48 kHz for phone calls</li><li>Bland Speech v3 ranked first of 13 AI voices on Audio Realism Bench on 4 August 2026</li></ul></div><div class="cmp-proscons-cons"><p><strong>Cons</strong></p><ul><li>More than a hobby project needs, because Bland Speech is built for production calls</li></ul></div></div>

### Fish Audio overview

Fish Audio, from Hanabi AI Inc., says it is trusted by 8M+ users and says its models run in production at HeyGen, Retell, and Sanas. Its recommended production model is s2.1-pro, with drama-3, a directable model launched on 13 October 2026, above it on price. Fish open-sourced the earlier S2 on 9 March 2026, and the Fish Audio Community License of 13 October 2026 makes commercial use free only while annual revenue is under USD 500,000.

Instant cloning takes a clip of at least 10 seconds, and Fish says a minute or two of clear, single-speaker speech improves fidelity. Professional Voice Clone lives in the web app, on a paid plan, needs at least 10 minutes of clean audio of one speaker, and asks the owner to confirm they own the voice or have the speaker's consent before release.

<div class="cmp-proscons" data-rival="fish-audio"><div class="cmp-proscons-pros"><p><strong>Pros</strong></p><ul><li>$15 per million UTF-8 bytes on s2.1-pro, with no subscription for API access</li><li>83 languages on s2.1-pro</li><li>Natural-language bracket tags that are not limited to a fixed set</li><li>Open S2 weights, and a self-hosted tier with VPC, on-prem, air-gapped, and sovereign cloud options.</li><li>HIPAA-aligned configurations with a BAA for qualifying healthcare workloads, and zero data retention on enterprise contracts</li></ul></div><div class="cmp-proscons-cons"><p><strong>Cons</strong></p><ul><li>Professional Voice Clone runs in the web app and needs a paid plan</li><li>The free plan is for personal, non-commercial use only</li><li>A SOC 2 audit that Fish says is underway, with the report to follow under NDA</li><li>Self-hosted deployment starts from a $10K platform fee and a $10K monthly minimum.</li><li>Community License terms that limit commercial use of the open weights above USD 500,000 in annual revenue</li></ul></div></div>

### Who owns a cloned voice, and who can use it?

On Bland Speech the answer is one organization. Clones are private to the customer's organization, and the customer confirms each voice is theirs to use. On Fish Audio the owner confirms permission before releasing a professional voice, and a released voice can earn a share when other creators pay to use it, which suits a creator who wants to earn from a voice and does not suit a brand with a spokesperson under contract. Fish's free plan allows personal, non-commercial projects only, and its paid plans allow commercial use of verified voices a user owns. More on law and consent sits in [is voice cloning legal for business](/blog/is-voice-cloning-legal).

## How much does a cloned voice cost to run on Bland Speech and Fish Audio?

Fish Audio's API costs less per character than Bland Speech on s2.1-pro, and its professional clones sit behind a paid app plan, while Bland Speech charges one standard rate.

### Bland Speech pricing

Bland Speech costs $0.04 per 1,000 characters, which is $40 per million, in the studio or the API, with no monthly plan. New accounts get 50,000 characters free, once, and credits never expire. A 10-minute narration costs about $0.36, and a million characters costs $40. Enterprise pricing is custom, set with Bland's sales team.

### Fish Audio pricing

Fish Audio's API is pay as you go, and Fish says there are no subscription fees or monthly minimums for API access. Text to speech is $15 per million UTF-8 bytes on s2.1-pro and s2-pro, $50 on drama-3, and $0 on a free s2.1-pro model with fair-use limits and no latency or data processing agreement guarantees.

The web app has its own plans. Free costs nothing and includes 8,000 credits a month, and Plus is £11.49 a month at list, with 250,000 credits and 1 professional voice slot. Pro is £76.99 a month at list, with 2,000,000 credits and 5 professional voice slots. Fish's self-hosted enterprise tier starts from a $10K one-time platform fee and a $10K monthly minimum with a 12-month commit.

A team that wants Fish Audio's lower rate takes on a byte-based meter and a separate app plan for professional voices.

## Which sounds more human, Bland Speech or Fish Audio?

Bland Speech v3 ranked first of 13 AI voices in Design Arena's Audio Realism Bench results of 4 August 2026, at 1365, and the human recording scored 1500. Fish Audio wasn't in those results.

Those scores come from each model's most conversational stock American-English voice, which measures how human a model reads and does not measure how closely a clone matches a person, as [Design Arena's published results](https://notes.designarena.ai/audio-realism-benchmark-measuring-realism-in-audio-models/) set out in the method. Fish Audio's drama-3 is built for voice direction and content creation, with plain-language instruction for sounds, pauses, and delivery limited to specific words, and a buyer who wants acted performance from a clone should listen to it.

### How does a cloned voice handle a phone call?

Phone networks carry narrow 8 kHz audio, so a clone should be judged after that squeeze. Bland Speech returns mu-law at 8 kHz or PCM from 8 to 48 kHz and publishes 199ms to first audio, as streamed. Fish Audio returns mp3, wav, pcm, or opus, says its latency is under 70 ms, and gives 100ms time to first audio for s2-pro.

## Why is Bland Speech the best Fish Audio alternative for voice cloning?

A business that wants a clone nobody else can license, on a phone line, chooses Bland Speech over Fish Audio: clones are private to the organization, output includes mu-law at 8 kHz, and Bland Speech v3 ranked first of 13 AI voices on 4 August 2026. On compliance, Bland is SOC 2 Type II audited and HIPAA compliant with a signed BAA, and it is FedRAMP 20x Class C certified at Moderate, granted 17 November 2026, and Bland says speech synthesis runs inside the FedRAMP boundary.

More comparisons sit in [Bland Speech vs Fish Audio](/compare/bland-speech-vs-fish-audio), the [Fish Audio alternatives](/blog/fish-audio-alternatives) list, and [how to move a cloned voice to Bland Speech](/blog/move-cloned-voice-to-bland-speech). Cloned voices are not portable between providers, so a move means re-cloning from the original recordings, with consent.

### How was this comparison made?

Bland Speech and Fish Audio were compared on clone types, ownership, plans, price, phone audio, and sound, using the plans, prices, formats, and licence terms each company publishes. Realism scores come from Design Arena's published results of 4 August 2026.

## Frequently asked questions

### Can Fish Audio clone a voice from 10 seconds of audio?

Yes. Fish Audio says to aim for at least 10 seconds per clip, and says a minute or two of clear, single-speaker speech improves fidelity. Bland Speech makes an instant clone from one clean sample of about 10 seconds.

### Is a Fish Audio professional clone available through the API?

No. Fish Audio's Professional Voice Clone runs in its web app only, on a paid plan with a professional voice slot. Plus has 1 slot and Pro has 5, and the Free plan has none.

### Are Bland Speech clones private?

Yes. Clones are private to the customer's organization, and the customer confirms every voice is theirs to use. On Fish Audio, a professional voice that its owner releases can earn a share when other creators pay to use it.

### Is Fish Audio cheaper than Bland Speech?

Per unit, yes. Fish Audio charges $15 per million UTF-8 bytes on s2.1-pro, and Bland Speech charges $40 per million characters. Bytes and characters match for plain English and diverge for other scripts, and Fish's professional clones need a paid app plan.

### Which audio formats does Fish Audio output?

Fish Audio's text to speech formats are mp3, wav, pcm, and opus. Bland Speech returns mu-law at 8 kHz for phone lines, as well as PCM from 8 to 48 kHz.

### Is Fish Audio open source?

Partly. Fish open-sourced its S2 model on 9 March 2026, and its Community License of 13 October 2026 allows free commercial use only while annual revenue is under USD 500,000. Bland Speech is a closed-source model Bland built itself.

### Does Bland Speech sound more human than Fish Audio?

Bland Speech v3 ranked first of 13 AI voices in Design Arena's Audio Realism Bench results of 4 August 2026, at 1365. Fish Audio wasn't in those results.
