What it does
Onepin is an AI voice production agent that sits on top of text-to-speech engines. You bring a script; Onepin cleans the text, routes each line to one of 30+ commercial TTS models picked per language from continuous benchmarks, scores every generated line for accuracy, naturalness, noise and pronunciation, and regenerates or patches lines that fail before you export the audio.
Official source 1 ↗ · Official source 2 ↗Features
- Multi-model TTS routing. One account reaches 30+ commercial TTS engines, including ElevenLabs, OpenAI, Microsoft, Google, AWS and Cartesia. Onepin benchmarks models across languages continuously and auto-picks the model for each line. Official source 1 ↗ · Official source 2 ↗
- Text normalizer. Before synthesis, the script is cleaned so numbers, dates, prices, addresses and abbreviations are expanded into how they should be spoken. Official source 1 ↗
- Pronunciation dictionary. A 4-million-word dictionary covers brand and product names, medical and drug terms, and hard personal names across languages. Official source 1 ↗
- Per-line validation scores. Every generated line is scored for word accuracy (WER), naturalness, noise or clarity, and pronunciation against a threshold you set. Official source 1 ↗ · Official source 2 ↗
- Automatic fixes. Lines that fail validation are regenerated, or only the misread word is fixed, using the same model with new settings. Official source 1 ↗
- Bring your own key. Studio and higher plans can run generation on your own TTS provider API keys, such as ElevenLabs, while Onepin handles routing, validation and fixing. Official source 1 ↗
- Commercial-use outputs and team pooling. Outputs are assigned to you and only engines that permit commercial use are routed to. Seats are free on every plan; Enterprise adds pooled credits across workspaces and a negotiated IP indemnity. Official source 1 ↗ · Official source 2 ↗
- Data-use and security controls. Onepin says it does not use customer content, including scripts and voice recordings, to train its own or third-party models, and that it is SOC 2 Type II compliant. Official source 1 ↗ · Official source 2 ↗
Pricing & free limits
1,000 one-time credits (about 15 minutes of validated audio), 1 workspace, 1 concurrent run, 1 GB storage, 10 languages and the full validation pipeline. No card required, no overage, and downloading audio to disk is not included. Official source 1 ↗
3,000 credits a month (about 100 minutes of validated audio), 3 concurrent runs, 100 GB storage, audio export, email support and overage at $0.010 per credit. Official source 1 ↗ · Official source 2 ↗
15,000 credits a month (about 9 hours of validated audio), 3 workspaces, 500 GB storage, bring your own key, email and Slack support, and overage at $0.008 per credit. Official source 1 ↗ · Official source 2 ↗
50,000 credits a month (about 30 hours of validated audio), 10 concurrent runs, 5 workspaces, 1 TB storage, the shared SOC 2 annual report and the lowest overage rate of $0.007 per credit. Official source 1 ↗ · Official source 2 ↗
Custom credit volume pooled across teams, unlimited seats and workspaces, negotiated IP indemnity, SAML SSO, custom data retention, DPAs and BAAs, a 99.9% uptime SLA and dedicated support. Official source 1 ↗
Billing details
TTS generation is included in credits rather than billed separately; Onepin passes generation through at cost. Official source 1 ↗
Unused credits do not roll over on self-serve plans; each billing cycle starts fresh. Official source 1 ↗
Past a plan's credits, paid plans meter usage as pay-as-you-go overage; Free has no overage. Official source 1 ↗
Annual billing saves 20%, and you can downgrade to Free or cancel at any time. Official source 1 ↗
Free access & limits
The Free plan includes 1,000 one-time credits, roughly 15 minutes of validated audio, with one workspace, one concurrent run and 1 GB storage. Saving generated audio to disk requires a paid plan; in-app playback is free. Official source 1 ↗
Best for & limitations
Producing audiobook, e-learning, gaming, advertising or film voiceover from a script, with mispronunciations caught before delivery. Official source 1 ↗
Generating multilingual product, support or healthcare narration where brand names and medical terms must be pronounced correctly. Official source 1 ↗
Adding routing and automatic quality checks on top of an existing ElevenLabs or other TTS provider account. Official source 1 ↗
- 10 languages. Ten languages are supported on every plan at the time of checking. Official source 1 ↗ · Official source 2 ↗
- No voice cloning or open-source models yet. Voice cloning and open-source TTS models are listed as coming soon; routing currently covers commercial models only. Official source 1 ↗
- Human review not yet available. Human validation is described as a coming-soon premium add-on; checks are automatic today. Official source 1 ↗
- GDPR compliance in progress. Onepin says it is working toward GDPR compliance. Official source 1 ↗
- Routed providers have their own terms. Third-party TTS providers operate under their own terms, and Onepin says it cannot guarantee every routed provider's data practices. Official source 1 ↗