# Fish Audio — Expressive text-to-speech, voice cloning and long-form studio with a credit-metered API. > Source: Dubsheet — https://dubsheet.com/fish-audio (structured, researched, re-verified) Fish Audio is a browser studio and API for generating speech from text with emotion tags, cloned voices and a community voice library. It also transcribes audio, changes voices, and assembles multi-speaker long-form projects in Story Studio. **Best for:** Creators narrating videos, audiobooks and character voices from their own cloned voice | Fact | Value | | --- | --- | | Website | https://fish.audio/ | | Category | Voice cloning, Courses and e-learning, Free text to speech, Text to speech and AI voice-over, YouTube and social video | | API | Yes | | Open source | No | | Price | Free | | Price checked | 28 October 2026 | Screenshot of Fish Audio: https://dubsheet.com/img/1920/screenshots/fish-audio.webp, captured 26 October 2026 **What we checked** - Pages last read: 28 October 2026 - Price on its site: Free plan (From its pricing page) ## Fish Audio in our research **Also named in** - [API text to speech with consent](https://dubsheet.com/collections/api-text-to-speech-with-consent) · Collection ## Fish Audio price and terms **Terms** - Free plan: Yes — Free Tier is $0/month with no credit card needed: 8,000 credits monthly, up to 7 minutes generation, up to 500 characters per generation, 3 public voice slots and standard generation speed; enhanced voice cloning and commercial use are marked as not included. - Sign up without sales: Yes **What needs a higher plan** - Pro: Cost for 10 hours a month (heavy use), Team seats - Plus: Professional cloning, Lip sync ## For AI agents - API: Yes Read on https://fish.audio/plan/, 28 October 2026. ## What we checked on Fish Audio ### What an hour of speech costs **Cost for 2 hours a month** · Yes — Two hours of speech needs 120 of the 200 minutes the $15/month Plus plan includes, and Plus is the cheapest tier whose feature list states commercial use is allowed. > Up to 200 minutes generation > — https://fish.audio/plan/ · Plan cards, Plus column, checked 28 October 2026 **Cost for 10 hours a month (heavy use)** · Yes — Ten hours is 600 minutes, over the Plus allowance, so it needs Pro at $100 a month month-to-month, which includes up to 1,620 minutes of generation. > Up to 1,620 minutes generation > — https://fish.audio/plan/ · Plan cards, Pro column, checked 28 October 2026 **Promotion and regular price** · Yes — promotion: the plan cards opened with a campaign banner halving the first month, and the annual toggle also shows a per-month figure, so two discounts stack on the headline price. > Limited time: 50% off your first month > — https://fish.audio/plan/ · Campaign banner above the plan cards, Monthly toggle state, checked 26 October 2026 **What the price counts** · Yes — Plans are metered in monthly credits: 8,000 on Free, 250,000 on Plus, 2,000,000 on Pro and 25,000,000 on Max, with the FAQ converting credits into minutes of speech. > Each minute of generation costs roughly 600 to 625 credits. > — https://fish.audio/plan/ · Frequently Asked Questions, "How many credits does each minute of generation cost?", checked 26 October 2026 **Voices that use more** · Yes — Newer and borrowed voices cost more: on the MCP surface another team's professional clone is charged double, and in the API the drama-3 model lists at more than triple s2.1-pro. > Speech and Story Studio audio in another team's professional voice clone cost double, and failed generations are refunded. > — https://docs.fish.audio/overview/mcp · Credits section, paragraph under the cost table, checked 26 October 2026 **Unused credits** · No — Monthly allowances reset with the billing cycle and anything left is lost, which the vendor states plainly and pairs with advice to spend the allowance inside the period. > Monthly quotas reset at the beginning of each billing cycle. Unused minutes do not roll over to the next month. > — https://fish.audio/plan/ · Frequently Asked Questions, "What happens to my unused minutes?", checked 26 October 2026 **Commercial use** · Partly — Publishing commercially starts one plan above the free one: the Free Tier card lists commercial use among the features it does not include and the vendor restricts free output to personal projects, while Plus at $15 a month lists commercial use allowed, limited to voices you own and have verified. > Premium subscribers can use verified voices (that you own) for commercial purposes. Free plan users can only use generated content for personal, non-commercial projects. > — https://fish.audio/plan/ · Frequently Asked Questions, "Can I use generated voices for commercial purposes?" answer; plan cards show "Commercial use" crossed out on Free Tier and "Commercial use allowed" on Plus, checked 28 October 2026 ### The voices **Stock voices** · Yes — The voice library is community-driven and the vendor advertises more than two million uploaded voices to pick from, alongside its own curated defaults. > The Fish Audio platform hosts over 2,000,000 voices, ideal for diverse scenarios from creative storytelling and dynamic advertisements to immersive audiobooks and beyond. > — https://fish.audio/ · "2,000,000+ voices, infinite possibilities" section, checked 26 October 2026 **Languages** · Yes — The vendor names eight languages for text to speech with native accents, though a homepage solutions tile claims 30+ languages, so the count depends which page you read. > Fish Audio supports multiple languages including English, Japanese, Korean, Chinese, French, German, Arabic, and Spanish. > — https://fish.audio/ · Frequently asked questions, "What languages does the best AI voice generator support for text to speech?", checked 26 October 2026 **Re-voice your recording** · Yes — A Voice Changer tool in the web app takes an uploaded recording and renders it in another voice, so your own delivery can be carried into a different voice. > Upload audio and transform it into another voice. > — https://docs.fish.audio/overview/platform · Create audio section, Voice Changer card, checked 26 October 2026 **Emotion, pace and emphasis** · Yes — Delivery is directed with bracketed emotion and pause tags in the text, and the API exposes prosody speed from 0.5 to 2.0 plus volume for finer control. > Adjust speech speed (0.5–2.0) and volume. > — https://docs.fish.audio/features/text-to-speech · Implementation details, "Speed & prosody", checked 26 October 2026 **Pronunciation fixes** · Yes — Reusable pronunciation dictionaries are built and published in the web app and then referenced from text to speech requests, so one fix applies across a whole project. > Create, preview, and publish reusable pronunciation rules. > — https://docs.fish.audio/overview/platform · Voices section, Pronunciation Dictionaries card, checked 26 October 2026 **Long scripts and chapters** · Yes — Story Studio is a project editor for long-form, multi-speaker audio, and its credit rule says blocks whose audio is current are free, so single blocks can be regenerated cheaply. > Assemble multi-speaker, long-form audio (audiobooks, dialogue, and narration) in a project editor. > — https://docs.fish.audio/overview/platform · Produce projects section, Story Studio card, checked 26 October 2026 **Audiobook files** · Partly — The vendor advertises chapter-level audiobook narration meeting Audible's ACX specs, but it documents no retail-ready export package or store delivery step, only audio files. > Publish-ready storytelling with lifelike pacing, emotion, and chapter-level control. Generate hours of audio that meets ACX/Audible specs without a recording booth. > — https://fish.audio/ · Use-case carousel, "Audiobook Narration" panel, checked 26 October 2026 **Several speakers** · Yes — Story Studio projects are explicitly multi-speaker, and the s2-pro and drama-3 models support multi-speaker dialogue inside a single generation. > previous-generation S2 model with multi-speaker and natural-language expression control > — https://docs.fish.audio/features/text-to-speech · Implementation details, Models list, s2-pro entry, checked 26 October 2026 **Audio you get** · Yes — Downloads cover MP3 with selectable bitrate, uncompressed WAV at a chosen sample rate, raw PCM and Opus, so quality can be matched to the delivery channel. > `mp3` (default), `wav`, `pcm`, `opus`. Set `format` (and optionally `mp3_bitrate`, `sample_rate`). > — https://docs.fish.audio/features/text-to-speech · Implementation details, "Output formats", checked 26 October 2026 ### Cloning a voice **Instant cloning** · Yes — A clone can be made from a very short sample with no training step, and the Free tier already carries three public voice slots, though it lacks the enhanced cloning paid plans get. > Instead of a saved `reference_id`, pass raw audio plus its transcript to clone a voice on the fly, with no training step. Best with a clean 10–30s sample. > — https://docs.fish.audio/features/text-to-speech · Implementation details, "Instant voice cloning", checked 26 October 2026 **Professional cloning** · Partly — Professional Voice Clone trains on at least ten minutes of recordings for studio quality, but it is excluded from Free and each paid plan grants a fixed number of slots, one on Plus. > A paid plan with a free PVC slot. Professional Voice Clone is not included in the Free plan. > — https://docs.fish.audio/overview/platform/create-voice/professional-voice-clone · Before you begin section, bullet list, checked 26 October 2026 **Consent of the voice’s owner** · Yes — For professional clones you record a spoken prompt that is matched against your uploads, or upload authorization documents for staff review, and you tick a permission declaration before release. > You must have permission to clone the voice: you own it, or you have the speaker's consent. You confirm this before you release the voice. > — https://docs.fish.audio/overview/platform/create-voice/professional-voice-clone · Before you begin, warning callout, checked 26 October 2026 **Your voice trains their models** · Yes — The terms authorize Fish Audio to train its models on your content and usage data, and the only published way out is the Enterprise plan's zero data retention, not a self-serve switch. > Usage Data and Content may be used to develop, train, or enhance artificial intelligence or machine learning models that are part of Fish.Audio's products and services > — https://fish.audio/terms/ · "Usage Date and Content" section, checked 26 October 2026 ### Dubbing and translation **Video dubbing** · No — Audio Translation transcribes and translates speech into text you read, not dubbed audio in the speaker's voice, and the product takes audio files rather than video. > We use automatic speech recognition (ASR) to transcribe your audio into text in its original language, then translate that text into your chosen target language. Upload an audio file or recording and get translated text back. > — https://fish.audio/translate-audio/ · Frequently asked questions, "How does AI audio translation work?", checked 26 October 2026 **Lip sync** · Partly — Lip-sync exists for avatars built from your images and videos against a Fish Audio voice, and needs a Plus plan or higher; it is not offered as lip matching for a translated dub. > Video (including lip sync and talking avatars) needs a Plus plan or higher, and Story Studio export needs a paid plan. > — https://docs.fish.audio/overview/mcp · Good to know section, bullet list, checked 26 October 2026 **Captions and subtitles** · Partly — Word-level timestamps and speaker turns come back from transcription and the vendor's cookbook turns them into SRT or VTT cues, but that is a developer recipe rather than a caption download in the studio. > Group word-level timestamps into SRT/VTT cues. > — https://docs.fish.audio/features/speech-to-text · When to use it, "Captions & subtitles" card, checked 26 October 2026 ### Working with it **Team seats** · Partly — Shared team workspaces start at Pro with three seats included and ten on Max, but no price per additional seat is published anywhere on the plan page. > 3 team seats included > — https://fish.audio/plan/ · Plan cards, Pro column feature list, checked 26 October 2026 **API** · Yes — A REST and WebSocket API with Python and JavaScript SDKs is sold pay-as-you-go; the docs say no subscription is required while the plan FAQ says premium subscribers get access, which conflicts. > The Fish Audio API uses pay-as-you-go pricing based on actual usage. There are no subscription fees or monthly minimums for API access. > — https://docs.fish.audio/developer-guide/models-pricing/pricing-and-rate-limits · API Pricing, opening paragraph, checked 26 October 2026 **API price** · Yes — The developer docs publish a public pay-as-you-go rate card with no subscription floor: text to speech is billed per million UTF-8 bytes, which the vendor equates to roughly 12 hours of speech, and transcription is billed per audio hour. > The Fish Audio API uses pay-as-you-go pricing based on actual usage. There are no subscription fees or monthly minimums for API access. > — https://docs.fish.audio/developer-guide/models-pricing/pricing-and-rate-limits · API Pricing section, opening paragraph above the Text-to-Speech models price table, checked 28 October 2026 **Platforms** · Yes — web: the whole product runs in the browser with no install, and every tool from cloning to Story Studio is reachable from the signed-in web app. > The Fish Audio web app gives you every capability without writing code: generate speech, clone voices, produce long-form projects, and manage your account. > — https://docs.fish.audio/overview/platform · Platform, opening paragraph, checked 26 October 2026 **Works inside** · Yes — An OAuth MCP server plugs the account into Claude, Claude Code, Cursor, Codex and Windsurf, and official Python and JavaScript SDKs cover custom pipelines; no Canva, Adobe or WordPress plugin is documented. > The Fish Audio MCP server gives AI agents direct access to your Fish Audio account: they can browse the voice library, clone a voice, generate speech, transcribe audio, generate or edit images and videos > — https://docs.fish.audio/overview/mcp · MCP Server, opening paragraph, checked 26 October 2026 ### Trying it, starting and getting help **Free plan or trial** · Yes — A standing free tier needs no card and gives 8,000 credits a month, about seven minutes of audio, capped at 500 characters per generation with three public voice slots. > 8,000 credits monthly Up to 7 minutes generation Up to 500 characters per generation 3 public voice slots > — https://fish.audio/plan/ · Plan cards, Free Tier column feature list, checked 26 October 2026 **What the free plan leaves on your audio** · Partly — Free output may not be monetized or used commercially, which the plan card and FAQ both state; whether that audio is watermarked or credited is not addressed anywhere. > Fish Audio's free plan is for personal use only. To monetize content or use voices commercially (YouTube, podcasts, business), upgrade to our paid plans for full commercial rights. > — https://fish.audio/ · Frequently asked questions, "Can I use free AI voice generator for commercial use and monetization?", checked 26 October 2026 **Still sold** · Yes — live: plans are on sale with self-serve upgrade buttons, a current launch promotion runs on the pricing page, and a new drama-3 model is being marketed on the homepage. > Unleash your creativity with unlimited generations > — https://fish.audio/plan/ · Choose Your Plan heading area, above the live Upgrade buttons, checked 26 October 2026 **Sign up without a sales call** · Yes — Plus, Pro and Max are bought from the pricing page with card payment through Stripe; only the Enterprise tier with its custom volume pricing routes through Contact Sales. > We accept all major credit cards (Visa, MasterCard, American Express), debit cards, and various regional payment methods through our secure payment processor. > — https://fish.audio/plan/ · Frequently Asked Questions, "What payment methods do you accept?", checked 26 October 2026 **Contract and cancelling** · Yes — Both monthly and yearly billing are sold and you can move between them whenever you like, with the change landing at the next cycle; the pages do not say whether canceling is self-serve. > Yes, you can switch between monthly and yearly billing at any time. When switching to yearly, you'll save 33% compared to monthly billing. Changes take effect at your next billing cycle. > — https://fish.audio/plan/ · Frequently Asked Questions, "Can I switch between monthly and yearly billing?", checked 26 October 2026 **Support** · Partly — Support is by email to support@fish.audio with a Discord community alongside it; no live chat and no hours or response times are published anywhere on the site or docs. > Support Email - Contact our support team > — https://docs.fish.audio/developer-guide/models-pricing/pricing-and-rate-limits · Support section, resource list, checked 26 October 2026 ## Fish Audio alternatives - [HeyGen](https://dubsheet.com/heygen) · Free plan, then $29 a month: AI avatar video, voice cloning and video translation into 175+ languages with lip sync. - [Murf AI](https://dubsheet.com/murf) · From $29 a month: Voice-over Studio with timeline video syncing, downloadable audio, dubbing, and a public API. - [Voicemaker](https://dubsheet.com/voicemaker) · Free plan, then $5 a month: Credit-metered AI text-to-speech studio with 2,000+ voices, voice cloning, speech-to-speech and SRT export. - [Narakeet](https://dubsheet.com/narakeet) · Free plan, then $6 per 30 minutes: Duration-priced browser voice-over and narrated-video maker with non-expiring capacity. - [TopMediai](https://dubsheet.com/topmediai) · From $35.99 a month: AI studio for text to speech, voice cloning and video dubbing on one shared credit balance. - [Typecast](https://dubsheet.com/typecast) · Free plan, then $5 a month: Expressive AI voice generator with emotion controls, voice cloning and a developer speech API. - [Akool](https://dubsheet.com/akool) · Free plan, then $12 a seat a month: AI video suite with video translation, lip sync, text to speech, voice cloning and a voice changer. - [Speechma](https://dubsheet.com/speechma) · Free plan, then $9 a month: Free text to speech and creator audio tools with no account or credit card. - [Amazon Polly](https://dubsheet.com/amazon-polly) · Free plan, then $4 per 1m characters: Pay-per-character cloud text to speech from AWS, with four voice engines and a browser console. All 9 publish a price. [All 9 alternatives](https://dubsheet.com/fish-audio/alternatives/) ## Sources and changes Last changed 28 October 2026 **Sources** (10) - https://fish.audio/plan/, read 28 October 2026 - https://fish.audio/, read 26 October 2026 - https://docs.fish.audio/features/text-to-speech, read 26 October 2026 - https://docs.fish.audio/overview/platform, read 26 October 2026 - https://docs.fish.audio/developer-guide/models-pricing/pricing-and-rate-limits, read 28 October 2026 - https://docs.fish.audio/overview/mcp, read 26 October 2026 - https://docs.fish.audio/overview/platform/create-voice/professional-voice-clone, read 26 October 2026 - https://docs.fish.audio/features/speech-to-text, read 26 October 2026 - https://fish.audio/terms/, read 26 October 2026 - https://fish.audio/translate-audio/, read 26 October 2026 **Updates** (2) - 28 October 2026: Dubbing features, Free tier, Plan limits, Plans - 26 October 2026: Dubbing features, Free tier, License terms, Plans ## Is this yours? Claim Fish Audio: https://dubsheet.com/claim/fish-audio Report an error if a fact is wrong: hello@dubsheet.com, subject "Report an error: Fish Audio (fish-audio)".