⚡️ Introducing DETECT-2B, the most advanced deepfake detection model

Q

Deploy Resemble on your own premise

Bring the power of Resemble’s cutting-edge voice AI to your own secure environment with our flexible on-premises solutions.

TRUSTED BY DEVELOPERS AT

Developers at Age of Learning trust Resemble AI
Developers at Red Games trust Resemble AI
Developers at Netflix trust Resemble AI
Developers at Humane trust Resemble AI

pip install ready.

Get started with Resemble’s voice AI capabilities in minutes using our convenient Python package. Perfect for developers who want to quickly experiment or incorporate voice features into existing applications.

    Easy Installation

    Install the Resemble package directly from your Python environment using familiar pip commands. No complex setup or additional tools required.

    Secure and Self-Contained

    The resemble-local package runs entirely on your own machines, keeping your voice data and processing fully isolated. No internet connection or external dependencies needed.

    Flexible Licensing

    Choose the subscription plan that fits your needs, from individual seats to site-wide licenses. Upgrade anytime as your usage grows, without any change to your code.

    Why customers choose Resemble AI when self-hosting

    At Resemble AI, we’re proud to be the go-to choice for businesses and individuals looking to leverage the power of AI voice cloning.

    9

    200ms to Response

    Convert text to high-quality speech in just 200 milliseconds. Enable fluid, interactive voice experiences without compromising on responsiveness.

    Speech-to-Speech

    Go beyond simple text-to-speech by converting input speech directly to a new target voice. Unlock new possibilities like real-time voice translation and voice style transfer.

    Effortless Scaling

    Support high-volume voice workloads without incremental costs. Our optimized on-prem architecture lets you scale up as much as you need using your existing infrastructure.

    f

    Flexible Deployment

    Choose the on-prem flavor that fits your stack. Integrate our Python package directly into your applications, or orchestrate voice microservices using our Kubernetes resources.

    Data Security and Privacy

    Keep your voice data fully in-house and under your governance. No need to send sensitive audio to the cloud or third party services. HIPAA and data residency compliant.

    Developer Friendly

    Resemble's APIs and SDKs make it simple for your developers to voice-enable your products. Go from idea to voice prototypes in record time using familiar tools and frameworks.

    How should you decide between deployment options?

    Alongside our cloud service, Resemble offers two flexible deployment options for on-premises voice AI: a Python package and a containerized Kubernetes deployment. Both provide the full capabilities of Resemble’s voice cloning, text-to-speech, and speech-to-speech synthesis, but are suited to different use cases and technical environments. 

    Ease of Setup
    Scalability
    Performance
    Latency
    Security
    Maintenance
    Ideal Use Cases
    Python Package (On-Prem)
    Simple pip install, minimal configuration
    Limited by local machine resources
    Dependent on local hardware
    Lowest latency for local applications
    Full data isolation, no external network access
    User responsible for package updates and hardware maintenance
    Batch processing, local integrations, offline environments
    Kubernetes (On-Prem)
    Requires Kubernetes cluster and container management skills
    Highly scalable across distributed nodes
    Can be optimized with dedicated GPU nodes
    Low latency within private network
    Isolated within private cluster, customizable security policies
    User manages Kubernetes cluster and container updates
    High-volume, distributed workloads within private infrastructure
    Cloud API
    No setup required, instantly available
    Seamlessly scales based on demand
    Consistently high performance backed by cloud infrastructure
    Dependent on network connection to cloud
    Secure cloud infrastructure, but requires data transfer
    Fully managed by Resemble, no maintenance required
    Rapid prototyping, web/mobile apps, unpredictable traffic

    Refined AI Voices with Professional Voice Cloning

    Resemble’s on-premises solutions bring the power of professional-grade voice cloning to your own secure infrastructure. Our advanced machine learning models allow you to create stunningly realistic AI voices that capture the nuance and expressiveness of human speech.

    Frequently Asked Questions

    K
    L
    What are the system requirements for running Resemble's on-prem package?

    Resemble's Python package requires a standard 64-bit Linux or macOS environment with Python 3.7+. We recommend at least 16GB of RAM and an NVIDIA GPU for optimal performance. Docker and Kubernetes are required for the containerized deployment option.

    K
    L
    Can I use Resemble's on-prem solution without an internet connection?

    Yes, once installed, the Resemble on-prem package can run entirely offline without any external network connectivity. It is fully self-contained and does not require an internet connection for voice synthesis or processing.

    K
    L
    What kind of support is available for on-prem customers? How do I get help with integration or troubleshooting?

    All on-prem licenses include access to our premium support channels, staffed by dedicated voice AI experts. You'll have direct access via phone, email, and chat to get help with installation, integration, and any technical issues. We also provide extensive developer documentation and code examples.

    K
    L
    Can I use my own custom voice data with Resemble's on-prem voice cloning?

    Absolutely. Resemble's on-prem solution allows you to train voice cloning models using your own proprietary audio data. You have full control over the input data and resulting voice models, which never leave your secure environment.

    K
    L
    How secure is Resemble's on-prem solution? Does it comply with HIPAA, GDPR, or other data privacy regulations?

    Security and privacy are central to Resemble's on-prem architecture. All voice data and processing is fully isolated on your own servers, with no exposure to external networks. Our on-prem deployment options support HIPAA, GDPR, and PIPEDA compliance when properly configured. Detailed audit logs and granular access controls are available.

    K
    L
    Can I run Resemble's on-prem solution in an air-gapped or offline environment?

    Yes, Resemble's on-prem offerings are designed to run in fully air-gapped, offline environments. The package includes all necessary models and dependencies and does not require any network connectivity once installed. Special licensing is available for offline deployments.

    K
    L
    How does Resemble's on-prem performance and quality compare to the cloud-based API?

    Resemble's on-prem solutions use the same underlying voice models and architecture as our cloud offerings, ensuring equivalent output quality. Performance is dependent on your own hardware, but our optimized runtimes enable low-latency synthesis even on a single GPU. Benchmarks are available on request.

    Ready to deploy?

    Learn more about Resemble AI