Fish Audio

Fish Audio combines text-to-speech, voice cloning, transcription and voice-agent tools for creators and developers who want several audio workflows in one service. Its adjustable speech controls and broad API options are compelling; choose carefully, because monthly minutes expire and the free tier is personal-use only.

Overview

Fish Audio spans voiceovers, audiobook narration, character voices and conversational chatbots. Alongside speech generation, it offers voice cloning, speech-to-text and voice agents, making it a stronger fit for people with varied audio needs than for someone seeking only a basic read-aloud tool.

Text-to-speech supports emotion and expression controls, real-time generation, and adjustments for speed, volume and model parameters. It automatically supports eight languages with native accents. The library of more than 2,000,000 voices provides extensive choice, while WAV, PCM, MP3 and Opus output covers several common production formats.

Key features

Speech generation and cloning: Delivery controls give creators ways to shape a voice beyond its words, and voice cloning adds another option for production. The free plan's 500-character generation cap is restrictive for longer passages; paid plans raise that ceiling to 15,000, 30,000 or, for Max, no published per-generation figure.

Developer access: The API supports generation, cloning and transcription through REST, WebSocket streaming, and official Python and TypeScript SDKs. This is a useful range for integrating audio into applications; it is less relevant to users who only need a standalone voiceover workflow.

Enterprise deployment: Organizations can choose VPC, on-premises, air-gapped or sovereign-cloud deployments. Default data stays in the United States, while self-hosting keeps deployment inside the customer's infrastructure. Fish Audio says its SOC 2 Type II audit is underway; qualifying healthcare workloads can use HIPAA-aligned configurations and BAAs, and enterprise contracts can enable Zero Data Retention.

Formats and connections: WAV, PCM, MP3 and Opus outputs suit different audio workflows. Enterprise integrations include Vapi, Twilio, Retell and workflow automation tools.

Pricing

The Free Tier costs nothing and provides 8,000 credits monthly, up to 7 minutes of generation, up to 500 characters per generation and three public voice slots. Generation runs at standard speed, and use is personal and non-commercial. It is a practical way to explore the service, but its short output allowance and commercial restriction rule it out for paid production.

Plus is $11 a month, billed $132 billed annually. It includes 250,000 monthly credits, up to 200 minutes of generation, up to 15,000 characters per generation, unlimited public and 10 private voice slots, and one professional voice slot. It suits an individual creator needing more room than the free tier, but not a team seeking multiple seats.

Pro is $75 a month, billed $900 billed annually. Its 2,000,000 monthly credits support up to 1,620 minutes and 30,000 characters per generation; the plan also includes three team seats, unlimited voice slots, five professional voice slots and a seven-day money-back guarantee. It is the clearest fit for a small team or heavier production schedule.

Max is $749 a month, billed $8988 billed annually. It raises the allowance to 25,000,000 credits and up to 6,250 minutes, with 10 team seats and 15 professional voice slots. That scale points to high-volume teams; it is a substantial step up from Pro for users who do not need its larger capacity.

Enterprise has custom pricing, with volume pricing, pay-as-you-go organization controls, Zero Data Retention and on-premises deployment. Enterprise support includes 24/7 production support, a technical account manager and a stated 99% uptime SLA. Unused monthly minutes do not roll over, so choose a quota that matches expected use rather than counting on a later billing cycle to absorb a quiet month.

Platforms

Fish Audio supports web, API, Linux, macOS, Windows and self-hosted deployments. That combination accommodates browser-based use, application integration and infrastructure-level deployment, though the appropriate choice depends on whether the priority is individual production or organizational control.

Who it's for

Fish Audio is best suited to creators who want controllable speech for narration or character work, developers adding speech capabilities through an API, and organizations that need deployment options or operational support. Its range is less persuasive for someone who needs only a small amount of non-commercial speech, or for teams that cannot use their monthly minutes before they expire.

Pros and cons

Pros: TTS controls, cloning, transcription and voice agents share one service, useful for mixed audio workflows.

Pros: REST, WebSocket, Python and TypeScript access give developers multiple integration routes.

Pros: Enterprise deployment spans VPC, on-premises, air-gapped and sovereign-cloud environments, with a 99% stated uptime SLA and 24/7 production support.

Cons: Free use is personal and non-commercial, with only 7 minutes monthly and a 500-character generation cap.

Cons: Unused monthly minutes expire, making the plans a weaker fit for irregular workloads.

Cons: Pro and Max carry much larger annual-billed commitments than Plus, so the higher quotas make sense only when the additional capacity and seats are needed.

Alternatives

For a lightweight free option with open-weight model access, Kokoro TTS offers an 82-million-parameter model under an Apache license; choose it instead if that model and its free pricing suit your requirements better than Fish Audio's broader paid plans. Google Cloud Text-to-Speech is an API-focused alternative with Standard voices priced at US$0.000004 per character (US$4 per 1 million characters) and 0 to 4 million characters free; it may suit readers who prefer character-based pricing.

Typecast offers a free plan with three projects, 3,000 lifetime credits, 1 GB media storage and 720p video exports, making it worth considering when those project and export limits fit the job. LOVO AI has a free plan limited to five projects and a Pro+ tier with 20 voice-generation hours per month and unlimited projects; pick it if those project terms match your workload. NaturalReader is another freemium option across web, desktop and mobile platforms, while Speechify offers a free tier with 10 voices and a 1.5x maximum speed.

Acapela Group and CereProc are alternatives for readers comparing paid speech services.

Browse AI voice generators, AI voice cloning tools or text-to-speech tools for more options in those categories.

Verdict

Choose Fish Audio if you need speech generation alongside cloning, transcription or voice-agent capabilities, especially when API access or enterprise deployment matters. Its breadth and configurable output are the main reasons to choose it; look elsewhere if your use is occasional, since monthly minutes expire, or if you need commercial use without moving beyond the non-commercial free tier.