How can I get access to try self-hosting Deepgram?

Contact us to talk to an AI expert about the benefits of self-hosting your voice AI solution.

Self-hosted voice AI

Q: How do Deepgram’s self-hosted compute requirements compare to other self-hosted solutions?

Deepgram’s world-class inference speed directly translates to more efficient compute costs. Self-hosting Deepgram can be 10-100x more compute efficient than other self-hosted options, including Whisper, which dramatically reduces infrastructure costs.

Q: How do I know how much compute to allocate for Deepgram?

Your Deepgram Account Representative will work closely with you to understand your use case and provide a customized suggestion. This makes it easy for you to serve your highest traffic peaks, as well as scale down during slow periods to save on hardware costs.

Q: I have regulatory and data privacy constraints that don’t allow a third-party hosted API to be approved at my company. Is Deepgram an option?

If our hosted/SaaS API is not an option for you, Deepgram’s self-hosted offering can likely meet your security requirements. When self-hosting Deepgram, no audio data, transcripts, or text input will ever be sent to Deepgram, so you can have peace of mind that you’re protecting your customers’ data.

Q: Do I need separate environments or virtual machines for each product, language, or feature?

No! Deepgram can be deployed with just two containers that together provide speech-to-text, text-to-speech, audio intelligence, and dozens of languages and features for each product.

Deepgram’s cutting-edge voice AI is available to self-host on your own infrastructure, both in the cloud or on-premises. Unlock a higher level of performance and privacy in speech-to-text, text-to-speech, and language understanding. Self-hosting gives you full authority over how voice capabilities are deployed in your applications.

Get a demo View enterprise solutions

Your domain, our intelligence

Bring our advanced voice AI capabilities into your own environment.

<200ms latency
The fastest real-time inference speeds, co-located with your application to eliminate network latency.
Data privacy
Your audio never leaves your environment. Protect your customer's data without sacrificing the quality of your voice integration.
Seamless integration
Easily incorporate into your existing infrastructure, with support for Kubernetes, Docker, Podman, and other leading container orchestrators.
Scale on demand
Powerful auto-scaling out-of-the-box to serve production-scale traffic patterns.
Multi-platform versatility
Comprehensive guides for major cloud providers and bare metal setups.
No feature limits
The same API and feature set as our hosted API, available for self-hosting.
Cost efficiency
Built-in down-scaling during off-peak hours trims your cloud bill without compromising on performance during high demand.
Regulatory compliance
Meet strict industry regulations and data residency requirements by keeping all processing within your controlled environment.
Complete control
Manage every aspect of your voice AI infrastructure, from deployment to customization, ensuring full alignment with your specific needs.

Enterprise cloud savings

Deepgram Enterprise agreements can be negotiated through the AWS or GCP marketplaces. This allows your Deepgram usage to contribute to your cloud provider's committed spend program, helping you meet your cloud budget goals and unlock substantial discounts.

Amazon Web Services

Private offers negotiated through our AWS Marketplace listing count towards the AWS Enterprise Discount Program (EDP).

Google Cloud Platform

Private offers negotiated through our GCP Marketplace listing count towards your Committed Use Discounts (CUDs).

FAQs

Deepgram’s world-class inference speed directly translates to more efficient compute costs. Self-hosting Deepgram can be 10-100x more compute efficient than other self-hosted options, including Whisper, which dramatically reduces infrastructure costs.

Your Deepgram Account Representative will work closely with you to understand your use case and provide a customized suggestion. This makes it easy for you to serve your highest traffic peaks, as well as scale down during slow periods to save on hardware costs.

If our hosted/SaaS API is not an option for you, Deepgram’s self-hosted offering can likely meet your security requirements. When self-hosting Deepgram, no audio data, transcripts, or text input will ever be sent to Deepgram, so you can have peace of mind that you’re protecting your customers’ data.

No! Deepgram can be deployed with just two containers that together provide speech-to-text, text-to-speech, audio intelligence, and dozens of languages and features for each product.

Purple paper texture fading to amber, with a carved stone figure at the right

Trusted by startups and enterprises

Discover the power of our product through real stories.

Explore customer stories

“Twilio’s flexible orchestration capabilities and global communications infrastructure combined with speech recognition powered by Deepgram APIs deliver seamless, low-latency, and human-like AI agent experiences that are powering today’s Voice AI renaissance.”
Andy O'Dower
VP of Product Management, Twilio
"Our voice capabilities powered by Deepgram APIs deliver the accuracy, low latency, and quality our customers need to build reliable, real-time voice agents."
Nikhil Gupta
Founder and CTO, Vapi
“They don't just chase generic benchmarks; they tailor their models to deliver on the metrics that really matter to us.”
Josh Levin
Head of Product, Cresta