Picovoice AI Frequently Asked Questions
Find answers to frequently asked questions on the Picovoice Platform, Console, and Pricing. For software-specific questions, please refer to the dedicated FAQs at the bottom of each product page:
- Leopard Speech-to-Text
- Cheetah Streaming Speech-to-Text
- Koala Noise Suppression
- Eagle Speaker Recognition
- Orca Text-to-Speech
- Octopus Speech-to-Index
- Porcupine Wake Word
- Rhino Speech-to-Intent
- Picovoice Platform
- Cobra Voice Activity Detection
Picovoice Technology
What’s Picovoice's business model?
Picovoice sells its proprietary voice AI technology to enable enterprises to build voice user interfaces, transcription, and search engines on a subscription basis. Picovoice’s subscription model:
- Offers access to support, updates, and upgrades during the engagement. Enterprises apply iterative and agile development principles and enjoy the advances in voice recognition and other software, such as operating systems, without worrying about the support and being stuck with old technology,
- Helps enterprises manage their working capital effectively,
- Automizes consumption tracking, resulting in efficiency gains and cost savings.
How does Picovoice process voice data?
Picovoice processes voice data locally on the device without sending audio data to any third party or a cloud provider. Thus voice experiences built with Picovoice don’t get affected by network latency, congestion, outages, and throttling.
How does Picovoice train voice AI models?
Picovoice researchers continuously improve techniques and frameworks used to train algorithms. Picovoice applies transfer learning, hardware-aware training, and neural compression principles, resulting in efficient and accurate models running across platforms and reducing time-to-market.
Should I provide recordings to train voice AI models?
Picovoice’s unique approach to speech recognition differentiates itself from other vendors. Anyone can fine-tune models on no-code Picovoice Console instantly or use pre-trained ones.
Does Picovoice collect user data?
Picovoice does not track, store or collect user data.
How can I decide which voice AI technology works best for my product?
The best way to choose which voice AI technology to use is to try and test it in a real-world environment. You can find a quick summary below, or read the short strategy guide on selecting the best voice technology.
- Convert audio and video files such as recordings of interviews, meetings, or calls, podcasts, and voicemails into text with Leopard Speech-to-Text
- Transcribe real-time conversations such as live events, conferences, and meetings, or enable note-taking and voice typing with Cheetah Streaming Speech-to-Text
- Transcribe real-time conversations such as live events, conferences, and meetings, or enable note-taking and voice typing with Cheetah Streaming Speech-to-Text
- Make audio and video libraries discoverable to search for keywords, including proper nouns, slang even without knowing the exact spelling with Octopus Speech-to-Index
- Achieve crisp and clear conversations by removing background noise and enhancing speech with Koala Noise Suppression
- Identify and verify speakers, and personalize experiences simply by recognizing the user’s voice with Eagle Speaker Recognition
- Convert written text into spoken audio output with Orca Text-to-Speech
- Detect wake words (Alexa), always listening commands (turn the lights on), and monitor conversations for specific keywords (product name) with Porcupine Wake Word
- Add custom voice commands to software (set the brightness at 60%), create voicebots and IVRs, and navigate in menus (2022 Hyundai IONIQ 5 AWD) with Rhino Speech-to-Intent
- Activate software when someone starts speaking (robocalls) and clean audio and video data (remove non-speech) with Cobra Voice Activity Detection
How can I get support while building with Picovoice?
Picovoice offers several types of support options, allowing developers and enterprises to choose what works the best for them.
- Consulting: Ideal for Enterprise Plan users and custom needs.
- Dedicated Support: Ideal for Developer and Enterprise Plan users interested in getting a dedicated and fast response without disclosing project details publicly.
- Enterprise Support Add-on: Ideal for Forever-Free Plan users who need direct access to the Picovoice team before commercially engaging with Picovoice.
- Jumpstart: Ideal for Forever-Free Plan users to get a head start instead of self-guided explorations.
- GitHub Community Support : Ideal for Forever-Free Plan users who want to report issues, engage with the community.
Which languages does Picovoice support?
Picovoice currently supports seventeen languages: English, Arabic, Dutch, Farsi, French, German, Hindi, Italian, Japanese, Korean, Mandarin, Polish, Portuguese, Russian, Spanish, Swedish, and Vietnamese. Please check the product page if you’re looking for engine-specific information. If you have an opportunity requiring another language, engage with Picovoice Consulting to get a custom model trained for you!
Does Picovoice technology work across various accents and dialects?
Yes, Picovoice technology works well across accents and dialects. The best way to learn about it is to test Picovoice engines with your dataset. Picovoice Free Plan allows enterprises to evaluate and get familiar with the technology before investing commercially. Picovoice also publishes open-source benchmarks for each engine. Please see the speech-to-text benchmark, phonetic search benchmark, wake word benchmark, natural language understanding benchmark, voice activity detection benchmark, and noise suppression benchmark.
Do Picovoice voice AI models work in noise and reverberation?
Yes, Picovoice software is resilient to noise and reverberation. However, the performance of voice AI software depends on many factors. For example, some use cases may benefit from Koala Noise Suppression. The best way to assess the performance of voice AI software in the presence of noise and reverberation is to test it. Developers can leverage Picovoice Free Plan to test engines.
Does Picovoice technology work in far-field applications?
Most likely. The performance of voice AI software depends on many factors, including distance, ambient noise level, reverberation (echo), quality of the microphone, and audio frontend used (if any). The best way to learn is to test the software in your environment.
Can I use Picovoice software for telephony applications?
Picovoice engines expect audio with a 16kHz sampling rate. PSTN networks usually sample at 8kHz. It is possible to upsample but the frequency content above 4kHz is gone, and performance will be suboptimal. It is possible to train acoustic models for telephony applications for enterprise customers. Engage with Picovoice Consulting to find the best solution that works for you.
My audio source is 48kHz/44.1kHz. Does Picovoice software support that?
Picovoice software expects a 16kHz sampling rate. You will need to downsample. Typically, operating systems or sound cards (Audio codecs) provide such functionality; otherwise, you will need to implement it.
Does Picovoice offer voice AI consulting services?
Picovoice offers Consulting Services to Enterprise Plan users. We understand some projects require engagements via professional services or non-recurring engineering projects. Thus, the Picovoice team and our network of subject-matter experts offer Picovoice Consulting to strategize or execute projects using Picovoice technology.
Picovoice Console Access and Pricing
How do I train voice AI models?
You can train voice AI models on the Picovoice Console . Picovoice Console is a no-code platform with a web-based type-and-train interface. You can create an account for the Picovoice Console account immediately and start building without engaging with the Picovoice team.
How many models can I train?
Free Plan users can train 3 (three) custom wake words with Porcupine Wake Word, 10 (ten) contexts with Rhino Speech-to-Intent and 10 Speech-to-Text (Leopard or Cheetah) models every 30 days.
Developer Plan users can train 10 (ten) custom wake words with Porcupine Wake Word, 100 (one hundred) contexts with Rhino Speech-to-Intent, and 100 Speech-to-Text (Leopard or Cheetah) models every 30 days.
Enterprise customers can increase the number of model allowances depending on their business requirements.
How does Picovoice billing and consumption work?
Picovoice tracks consumption via the volume of data processed or the number of active users in the last 30 days.
How do you calculate the volume of data processed in the last 30 days?
The volume of data processed in the last 30 days is calculated by the amount of data processed by Leopard Speech-to-Text, Cheetah Streaming Speech-to-Text, Koala Noise Suppression, Eagle Speaker Recognition, Orca Text-to-Speech, and Octopus Speech-to-Index. While the tracking is done in seconds, usage is shown in hours under “Monthly Usage Hours” for simplicity. Picovoice Console shows account limits and usage in real time. An account owner can use one engine or all across an unlimited number of users.
How do you calculate active users in the last 30 days?
The number of active users in the last 30 days is calculated using the number of “things” that activate Porcupine Wake Word, Rhino Speech-to-Intent, and Cobra Voice Activity Detection. The “thing” can be anything like a mobile app, web browser, or an IoT device... A user can use only one or all three engines and process voice interactions with no limits. If a “thing” stops using the voice features enabled by Porcupine Wake Word, Rhino Speech-to-Intent, or Cobra Voice Activity Detection, it will not be an active user after 30 days. Picovoice Console shows account limits and usage in real time.
Picovoice does not track end users. Even if multiple people activate a “thing,” Picovoice counts it as one user. Similarly, even if the same person uses different “things,” each thing will be counted as a unique user.
Can I reset my AccessKey on Picovoice Console?
No, neither you nor Picovoice can reset your AccessKey. Please do not share it with 3rd parties.
Can I reset my allowances without waiting for 30 days?
No, you cannot reset your allowances. You shall wait until they reset.
Can Picovoice reset my allowances?
No, Picovoice cannot reset your allowances.
Can I create multiple accounts till my allowances reset?
No, creating multiple accounts for the same project violates the Terms of Use. If you need more allowances, we encourage you to upgrade your account to the Developer Plan. Picovoice does not tolerate any behavior that jeopardizes its success and is unfair to
- Free Plan users who stay compliant
- paying customers
- Picovoice team members who work hard to build things you want to use more
In case of violation, you may be asked to pay additional fees, and Picovoice may refuse to serve you. For more information, please refer to the Terms of Use.