Which languages are supported?

10 languages are supported on all plans, with TTS auto-routing picking the best model per language.

Do I own the generated audio?

Yes. Every output is assigned to you, and Onepin only routes to engines that permit commercial use of their outputs. Enterprise adds a negotiated IP indemnity.

Does it offer enterprise security and compliance?

Enterprise includes SOC 2 annual report, SAML SSO, DPAs & BAAs, custom data retention policies, private workspaces, and a 99.9% uptime SLA with service credits.

Can I integrate Onepin into my own pipeline?

Yes. API, CLI, SDK, Agent, and MCP access are included on every plan, and Studio and above support BYOK so generation runs on your existing TTS provider accounts.

Onepin — AI voice production agent. Human voice, right words.

Onepin is an AI voice production agent that turns scripts into production-ready speech. It routes your script line-by-line across 30+ TTS models (ElevenLabs, OpenAI, Microsoft, Google, AWS, Cartesia, Murf AI, Deepgram, MiniMax, Inworld, and 20+ more), cleans text before synthesis, corrects pronunciations against a 4M-word dictionary, and validates every generated line against four scores (naturalness, word accuracy, clarity, pronunciation) before auto-fixing any line that misses the bar.

Key Features

Multi-model TTS routing

Access 30+ text-to-speech model APIs behind one account, routed line-by-line, with auto-selection of the best model per language based on continuous benchmarking.

Text cleaning and normalization

Fixes addresses, prices, dates, abbreviations, and other common TTS misread causes before synthesis, in every supported language.

Pronunciation correction

A 4M-word dictionary teaches models how to say brand names (Versace, Hermès), medical and drug terms (semaglutide), and hard names (Siobhán, Xóchitl) correctly.

Per-line validation scoring

Every generated line is scored on naturalness, word accuracy, clarity, and pronunciation against a quality bar you set, consistently across languages.

Automatic fixes

Lines that miss the bar are regenerated with the same model under new settings, or just the missed word is fixed, without manual rework.

Output ownership and commercial use

Every generated output is assigned to you and cleared for commercial use; Enterprise adds a negotiated IP indemnity.

API, CLI, SDK, Agent, and MCP access

Integrate the validated pipeline into your own tooling or agent workflows, plus in-app playback.

Enterprise procurement readiness

SOC 2 Type 2, SAML SSO, private workspaces, custom data retention, DPAs & BAAs, and a 99.9% uptime SLA with service credits.

How It Works

Paste your script and set the quality bar

Upload or paste your script and pick a quality bar for the validators.

Step 1: Clean

The normalizer fixes abbreviations, numbers, and other common TTS misread causes before synthesis. This addresses about half of voice errors.

Step 2: Pronounce

The phonemizer applies the 4M-word dictionary so brand names, drug names, and hard names are pronounced correctly in any language.

Step 3: Validate

Every generated line is scored on naturalness, word accuracy, clarity, and pronunciation against your bar, with the same standard across languages.

Step 4: Fix

Lines that miss the bar are regenerated automatically with the same model under new settings, or the single missed word is fixed. You get the best take without manual rework.

Download or integrate

Pros & Cons

Pros

Routes across 30+ TTS engines behind one account, so a provider price hike, deprecation, or shutdown doesn't force a pipeline rebuild
Automates the full quality workflow — normalization, pronunciation, validation, and auto-fix — rather than just raw synthesis
Per-line scoring on four metrics with a consistent bar across all languages
Outputs are assigned to you and cleared for commercial use on every plan
Free unlimited seats on all plans, with pooled credits across workspaces at Enterprise
Procurement-ready at Enterprise: SOC 2 Type 2, SAML SSO, DPAs & BAAs, custom retention, and a 99.9% uptime SLA with service credits

Who It's For

Best for

Solo creators, VTubers, YouTubers, and podcasters shipping 1–2 shows
Game studios producing localized dialogue
Audiobook and film teams that need consistent voice quality across languages
Advertising, e-learning, and product teams that need every line to clear a fixed quality bar
Production teams and studios that need procurement-grade security, output ownership, and commercial-use clearance

Not ideal for

Projects needing more than the 10 currently supported languages

Pricing

Freemium, with a free tier and four paid tiers (cheapest $16/month billed yearly). Plans scale credits, storage, concurrent pipeline runs, workspaces, and support. Overage is $0.010/credit on Creator, $0.008/credit on Studio, and $0.007/credit on Scale; Enterprise pricing is custom. Yearly billing saves 20%.

Free

$0/month

1,000 one-time credits, ~15 min validated audio, 1 concurrent pipeline run, 10 languages, TTS auto-route, mispronunciation fixes, 1 workspace, 1GB storage, 1-year file retention, API/CLI/SDK and Agent & MCP, in-app playback (no downloads),

Creator

$16/month billed yearly (save 20%)

Everything in Free plus 3,000 credits/month (~100 min validated audio), 3 concurrent pipeline runs, 100GB storage, downloads & exports, email support, 3-year file retention, overage $0.010/credit.

Studio

$80/month billed yearly (save 20%)

Everything in Creator plus 15,000 credits/month (~9 hrs validated audio), 5 concurrent pipeline runs, 3 workspaces, 500GB storage, BYOK, email + Slack support, 5-year file retention, overage $0.008/credit.

Scale

$240/month billed yearly (save 20%)

Everything in Studio plus 50,000 credits/month, 10 concurrent pipeline runs, 5 workspaces, 1TB storage, lowest overage rate ($0.007/credit).

Enterprise

Custom (contact sales)

Everything in Scale plus output ownership assigned to you, commercial use on every model, negotiated IP indemnity, custom credit volume, unlimited seats & workspaces, pooled credits across workspaces, SAML SSO, custom data retention, DPAs &

FAQ

What does Onepin do?

It's an AI voice production agent that turns scripts into production-ready voice. It cleans text, fixes pronunciations, routes your script across 30+ TTS models, validates every line on four quality scores, and auto-fixes anything that misses the bar.

Is Onepin free to use?

Yes. The Free plan includes 1,000 one-time credits (~15 minutes of validated audio), 1 pipeline run, 10 languages, mispronunciation fixes, API/CLI/SDK and Agent & MCP access, and in-app playback. Downloads and exports start on the Creator plan.

How much does Onepin cost?

Creator is $16/month billed yearly, Studio is $80/month billed yearly, and Scale is $240/month billed yearly (each saving 20% vs monthly). Enterprise is custom-priced. Overage rates are $0.010/credit on Creator, $0.008/credit on Studio, and $0.007/credit on Scale.

Which TTS providers does Onepin support?

Onepin