⚡Stack Radar

Voice Cloning · Reviewed 26 Oct 2026

Gradium

Real-time voice AI API for building voice agents.

from $12 · free tier · gradium.ai ↗

Paid against free › Voice Cloning › Gradium

Gradium at a glance
55.3KEstimated total trafficModeled monthly visits25Ranking keywordsPublic search sample98/100Technical health15 pages checked0Referring domainsLive backlink summary

Overview

01

Summary

The useful answer before the supporting detail.

Gradium is a real-time voice AI API for building voice agents, offering text-to-speech, speech-to-text, voice cloning, and voice design S1. It is used by developers and enterprises to create conversational AI with low latency and scalable deployment options S1. The service provides a free tier and credit-based plans, with deployment from cloud to on-premises S1 S31.

02

Product identity and positioning

Name, category, positioning, audience, and first-party claims.

  • Identity: Gradium is a real-time voice AI API for voice agents S1.
  • Core purpose: It solves latency, reliability, and scale challenges in real-time voice conversations S1.
  • Intended users: Developers, builders, researchers, and creators building voice agents S1.
  • Product scope: A voice AI API platform, not a full agent framework; it provides TTS, STT, and voice cloning components S1.
  • Positioning: Gradium positions itself as a production-grade API for handling hard real-world voice problems S1.
  • Claimed distinction: It claims low-latency streaming with flat latency as concurrency climbs S1.
03

Use cases and audiences

Who the product serves and the jobs it is designed to handle.

AudiencePractical use case
DevelopersBuild voice agents with a single-prompt SDK and streaming APIs S1.
EnterprisesDeploy voice agents in production with on-prem or dedicated instances for data sovereignty S1.
ResearchersUse peer-reviewed research translated into production-grade APIs S1.
StartupsApply for the startup program to get credits and support for building voice experiences S69.

Features

09

Features and capabilities

Core functions and evidence-backed product capabilities.

CapabilityWhat it does
Text-to-SpeechSynthesizes natural speech from text via API, with 1 credit per character S31.
Speech-to-TextTranscribes audio in real time, with 3 credits per second S31.
Voice CloningCreates instant voice clones from ~10 seconds of audio, available immediately S39 S152.
Voice DesignAllows designing custom voices from text descriptions S31.
Streaming WebSocket APIProvides bidirectional streaming for low-latency voice interactions S1.
On-device TTSPhonon model runs locally on devices with as little as 64.5 MB S1.
10

Integrations and ecosystem

Connected products, exports, APIs, and workflow compatibility.

Integration or interfaceTypeWhat it enables
LiveKitIntegrationOfficial integration for orchestration frameworks S92.
PipecatIntegrationOfficial integration for orchestration frameworks S92.
VapiPlatformBundled platform for phone agents S92.
Retell AIPlatformBundled platform for phone call automation S92.
WebSocket APIAPIBidirectional streaming for real-time voice S1.
Python SDKSDKOfficial SDK for Python developers S52.
Rust SDKSDKOfficial SDK for Rust developers S52.

Pricing

11

Pricing and plans

Published plans, free access, trials, credits, and pricing limitations.

Plan or offerVerified price and billingIncludedLimits
Free$0 monthly45,000 credits/month, 5 custom voicesNo commercial use, no pro clones S31
XS$12/month billed yearly (one month free)225,000 credits/month, 1,000 custom voicesCommercial use S31
S$39/month billed yearly (one month free)900,000 credits/month, 1,000 custom voicesZero data retention S31
M$312/month billed yearly (one month free)9,000,000 credits/month, 5 pro voice clonesS31
L$1,480/month billed yearly (one month free)45,000,000 credits/month, 20 pro voice clonesS31
EnterpriseContact salesUnlimited usage, on-prem deploymentCustom concurrency S31

Competition

12

Competitive landscape

Domains competing for the same organic search demand.

Sampled competitors

CompetitorFirst-page appearancesSampled queriesAverage positionQueries observed
fantasynamegenerators.com111robotic sounding names
fr.linkedin.com112gradium ai
reddit.com112robotic sounding names
techcrunch.com113gradium ai
wikihow.com113robotic sounding names
voiceaispace.com114gradium ai
pinterest.com114robotic sounding names
github.com115gradium ai
quora.com115robotic sounding names
ai.engineer116gradium ai

Website intelligence — Collected evidence

Customers, testimonials, and first-party review claims

No supporting public-page evidence was collected for this category.

Independent review sources discovered

No links to supported independent review platforms were discovered on the collected pages.

13

AI visibility

Measured ChatGPT and Google AI mentions with cited sources.

AI-assistant referral traffic

  • Estimated current generative-AI referral share: 1.3%
MonthSource
—All AI assistants

Website intelligence — Collected evidence

Site-stated AI capabilities

Page typeEvidenceSource
identityGet in touch with the Gradium team: enterprise voice AI and custom pricing, partnerships, support, press or anything else.https://gradium.ai/contact
identityJoin Gradium and help build the future of voice AI. Explore open roles in research, engineering, and product for text to speech, speech to text, and voice cloning.https://gradium.ai/about/careers
identityWe're always looking for talented individuals who are passionate about voice AI. Send us your resume and tell us how you can contribute.https://gradium.ai/about/careers
pricingGradium extends its seed funding to $100 million, welcoming new investors including NVIDIA, and opens a San Francisco Bay Area office to scale its real-time voice AI.https://gradium.ai/pricing
pricingStart freehttps://gradium.ai/pricing
pricingGradium extends its seed funding to $100 million, welcoming new investors including NVIDIA, and opens a San Francisco Bay Area office to scale its real-time voice AI.https://gradium.ai/pricing-faq
integrationsTalk to usSign inhttps://gradium.ai/
integrationsGradium extends its seed funding to $100 million, welcoming new investors including NVIDIA, and opens a San Francisco Bay Area office to scale its real-time voice AI.https://gradium.ai/
integrationsDocsStart building in minutesRelease NotesSee what's new in GradiumStatusMonitor system performanceTrustSecurity and compliancehttps://gradium.ai/
integrationsTry nowTalk to ushttps://gradium.ai/
integrationsFull suite of voice AI models to power your agentshttps://gradium.ai/
integrationsTalk to usSign inhttps://gradium.ai/app-submission
integrationsGradium extends its seed funding to $100 million, welcoming new investors including NVIDIA, and opens a San Francisco Bay Area office to scale its real-time voice AI.https://gradium.ai/app-submission
integrationsDocsStart building in minutesRelease NotesSee what's new in GradiumStatusMonitor system performanceTrustSecurity and compliancehttps://gradium.ai/app-submission
changesTalk to usSign inhttps://gradium.ai/blog/gradium-releases-phonon
changesGradium extends its seed funding to $100 million, welcoming new investors including NVIDIA, and opens a San Francisco Bay Area office to scale its real-time voice AI.https://gradium.ai/blog/gradium-releases-phonon
changesDocsStart building in minutesRelease NotesSee what's new in GradiumStatusMonitor system performanceTrustSecurity and compliancehttps://gradium.ai/blog/gradium-releases-phonon
changes[](https://x.com/intent/post?text=Gradium%20released%20Phonon%2C%20an%20open-source%20Text-to-Speech%20model%20that%20runs%20entirely%20on%20device%3A%2031m%20parameters%2C%20five%20languages%2C%20Apache%202.0.&url=https%3A%2F%2Fgradium.ai%2Fblog%2Fgradium-releases-phonon "Post on X")[](https://www.linkedin.com/sharing/share-offsite/?url=https%3A%2F%2Fgradium.ai%2Fblog%2Fgradium-releases-phonon "Share on LinkedIn")https://gradium.ai/blog/gradium-releases-phonon
changesBERLIN, NVIDIA GTC, October 21, 2026. Gradium, the voice AI company founded by the cofounders of Kyutai research lab, today released Phonon, the best performing on-device text-to-speech system. Phonon runs entirely on the user's hardware: speech is generated locally, with no network round-trip and no cloud dependency, so text and audio stay on the device.https://gradium.ai/blog/gradium-releases-phonon
changesPhonon is available today. The multilingual model weights are on Hugging Face at [](https://huggingface.co/gr4d/phonon-9444a467.200)0 and the inference code is on GitHub at [](https://github.com/gradium-ai/xn-ptts)1, under the Apache 2.0 licence. Voice cloning and voice design are available through the Gradium API at [](https://gradium.ai)2. Launch partners include Hugging Face, Arm, Qualcomm, Appning by FORVIA, LM Studio, Ollama, MLX Audio.https://gradium.ai/blog/gradium-releases-phonon

These are first-party product claims, not independent measurements of AI-search visibility.

14

Marketing opportunities

Actions calculated directly from the collected evidence.

OpportunityRecommended actionEvidence
Striking-distance keywordMove “robotic sounding names” from position 5 toward the top three.Volume 50 · https://gradium.ai/content/how-to-make-ai-voices-sound-more-natural
Important pageStrengthen the page already earning search visibility before creating an unrelated page.https://gradium.ai/content/how-to-make-ai-voices-sound-more-natural · 1 sampled keywords · best position 5
Competitor gapReview the sampled queries where fantasynamegenerators.com appears.1 first-page appearances across robotic sounding names

Website intelligence — Collected evidence

Website evidence gaps

  • features: no usable public page was collected; verify whether the site publishes this evidence.
  • customers: no usable public page was collected; verify whether the site publishes this evidence.
  • reviews: no usable public page was collected; verify whether the site publishes this evidence.
  • technical: no usable public page was collected; verify whether the site publishes this evidence.

Trust and proof

15

Customers, proof, and reviews

Published proof, customer signals, and review evidence.

Independent reviews

No citation-grounded independent review links were found.

Customer proof

No named customers, testimonials, or customer case studies were found.

Independent mentions

  • Real-Time Voice AI API for Voice Agents / Gradium — Real-Time Voice AI API for Voice Agents / Gradium ... Models Text-to-SpeechTurn text into natural sounding voiceSpeech-to-TextTranscribe audio into text instantlyLive TranslationConvert voice to voice in real timeVoice CloningInstance or pro voice cloneVoice DesignA brand-new voice from a text promptPhononOur open, on-device TTS model Showcase Agent DemoTest our voice agent liveGradbotBuild voice agents in a single promptBuildFrom idea to prototype in minutesCommunityBuilders, researchers and cr
  • About Gradium: The Team Building Advanced Voice AI / Gradium — Gradium builds the technological backbone, models and infrastructure, to support all voice applications. We work at the finest level of details to design natural, expressive, real-time voice interactions at scale.
  • Gradium API Documentation - Gradium API — Build voice-powered applications with the Gradium API. Our TTS and STT models deliver natural-sounding output with best-in-class accuracy, low latency, semantic VAD for turn-taking, and adaptive delay controls for realtime agents.
  • Gradium: Solving voice / Gradium — Today we’re excited to launch Gradium, the core engine powering the next generation of voice products and interactions. Drawing on more than a decade of frontier research, our mission is simple and ambitious: to provide the AI foundation that makes natural, real-time voice the default interface between people and machines. To achieve our ambition, we raised 70M$ in a seed round led by FirstMark Capital and Eurazeo with participation from DST Global Partners, Eric Schmidt, Xavier Niel, Rodolphe S
  • Pricing: Voice AI Plans and Credits / Gradium — ### XS $12/mo 1 month free 225k credits/mo Everything in Free, ... - Commercial use - 1,000 voices - EU or US residency ... $39/mo 1 month free 900k credits/mo Everything in XS, plus ... M $312/mo 1 month free 9M credits/mo Everything in
  • Voice AI Demo: Try Text to Speech and Speech to Text / Gradium — speech - Expressive voices and robust pronunciation - Instant and Pro voice cloning - Streaming inference for real-time applications - 48kHz audio output - ... audio alignment - Multilingual support: English, French, Spanish, German, and Portuguese
  • Gradium Extends Funding to $100 Million and Expands to Silicon Valley / Gradium — researchers behind Kyutai, the AI research lab behind pioneering real-time speech systems, Gradium was created to bring years of frontier AI research into production. Founded by Neil Zeghidour, Laurent Mazaré, Olivier
  • Blog: Voice AI Research, Engineering, and Product Updates / Gradium — Gradium releases Phonon, an open-source text-to-speech model that runs on any device Trained on NVIDIA GPUs, Phonon generates natural speech locally in as little as 64.5 MB, on phones, laptops, embedded hardware and in the browser.

External links were grounded in one live web search. Search snippets are leads, not independently scraped page content; rejected links remain available in Review evidence.

16

Company, security, and trust

Ownership, policies, security signals, and public trust evidence.

SignalFindingSource
Company and ownershipTalk to our team →source
Contact and supportPlans, credits and billing. For anything else, contact us.source
Legal and privacyWe’ll bill you monthly or yearly, starting from the day you subscribe. You can cancel your subscription at any time.source
Security and complianceSOC 2 Type IIIndependently audited controlssource
Uptime and status pageAll deployment models support the same API surface, the same models, and the same voice cloning capabilities. Enterprise plans include unlimited concurrent requests, multi-region deployment, auto-scaling, 99.9% uptime SLsource
Domain registered2025-03-04—
Domain expires2029-03-04—

Every trust signal above is backed by a collected page or a registration record.

17

Documentation and product updates

Help resources, product guidance, and visible changes over time.

Resource or updateTypeWhat is availableDate or freshness
Product FAQFAQAnswers common product, workflow, pricing, privacy, ownership, API, and team questions. (source)Not dated
How to Compare TTS Pricing Across Providers in 2026 / GradiumGuideGuides, comparisons, pricing breakdowns, use cases, and how-to articles. (source)Not dated
Gradium TTS in Pipecat: Setup and Integration Guide / GradiumGuideGradium TTS integration for Pipecat: install GradiumTTSService, configure voice and model settings, and stream low-latency speech in your voice agent. (source)Not dated
How to Add Text-to-Speech to an App or Website / GradiumGuideGuides, comparisons, pricing breakdowns, use cases, and how-to articles. (source)Not dated
How to Choose a TTS API: The Five Criteria / GradiumGuideGuides, comparisons, pricing breakdowns, use cases, and how-to articles. (source)Not dated
Real-Time Voice AI API for Voice Agents / GradiumChangelogPublic product change history. (source)Not dated
Best ElevenLabs Alternatives 2026: Quality vs Price / GradiumProduct updateProduct update is publicly available. (source)May 2026

Not found in the collected public sources: public roadmap, status page, separate help center or knowledge base.

Technical

18

Technical and delivery signals

Delivery model, platforms, API, data ownership, reliability claims and release activity.

SignalFindingSource
Delivery model (cloud, self-hosted, desktop, browser extension)Offers cloud API, dedicated instances, self-hosted, on-prem, and on-device deployment S1.S1
Platforms and appsWeb-based API with SDKs for Python and Rust S52.S52
API and extensibility (API, SDK, webhooks, MCP, plugins)Provides WebSocket streaming API and official SDKs S1 S52.S1
Data ownership and exportAudio data provided is used strictly to train your specific voice model or generate your clone S39.S39
Performance and reliability claimsLatency stays flat as concurrency climbs S1.S1
Release activityPhonon, an open-source TTS model, was released S1.S1
19

Measured infrastructure

Domain, certificate, and hosting signals, measured directly.

SignalFinding
Domain registered2025-03-04 — Key-Systems GmbH
Domain expires2029-03-04
First archived2025-12-02 — web archive
TLS certificateLet's Encrypt, valid, expires 2026-12-23
Security headersHSTS, X-Content-Type-Options, X-Frame-Options, Referrer-Policy
Headers not servedCSP
Mail handled bysmtp.google.com
Nameserversns-1488.awsdns-58.org, ns-1790.awsdns-31.co.uk

Not measurable: rdap: exceeded the overall budget; wayback: exceeded the overall budget; tls: exceeded the overall budget; dns: exceeded the overall budget; headers: exceeded the overall budget. A gap here means the lookup was unavailable, not that the finding is negative.

20

Technical site audit

A bounded multi-page crawl with page-level issues and URLs.

Scout — Bounded technical crawl

  • Pages crawled: 15 of 15 maximum
  • Technical health: 98/100
  • Critical issues: 0
  • Warnings: 5
  • Notices: 5

Issues

SeverityIssueValueAffected URL
warningPage should have one primary H10https://gradium.ai/
noticeMeta description length should be reviewed186https://gradium.ai/phonon
warningPage should have one primary H10https://gradium.ai/phonon
warningTitle length should be reviewed67https://gradium.ai/blog
noticePage is marked noindex—https://gradium.ai/contact
noticeLow visible word count53https://gradium.ai/contact
warningTitle length should be reviewed79https://gradium.ai/blog/gradium-100-million-funding-nvidia
noticeMeta description length should be reviewed182https://gradium.ai/voice-design
warningPage should have one primary H10https://gradium.ai/demo
noticeMeta description length should be reviewed206https://gradium.ai/gradbot

Crawled pages

StatusTitleH1sWordsResponseURL
200Real-Time Voice AI API for Voice Agents \Gradium0522618 ms
200Phonon: Open On-Device Text to Speech from Gradium \Gradium01703746 ms
200Pricing: Voice AI Plans and Credits \Gradium11353159 ms
200Blog: Voice AI Research, Engineering, and Product Updates \Gradium11189807 ms
200Talk to us \Gradium153637 ms
200Gradium Extends Funding to $100 Million and Expands to Silicon Valley \Gradium11567324 ms
200Text to Speech : Real-Time, Expressive Voices \Gradium11684172 ms
200Speech to Text : Real-Time Transcription API \Gradium11778316 ms
200Live Translation : Real-Time Speech Translation API \Gradium11622213 ms
200Voice Cloning : Instant and Pro Voice Clones \Gradium11682164 ms
200Voice Design: A Brand-New Voice From a Text Prompt \Gradium12605177 ms
200Voice AI Demo: Try Text to Speech and Speech to Text \Gradium0763171 ms
200Gradbot: Open Source Voice Agent Framework \Gradium11353181 ms
200Build with Gradium \Gradium11107166 ms
200Community \Gradium11103159 ms

This is a bounded Scout crawl. Core Web Vitals and Lighthouse scores are not claimed unless a dedicated performance audit is connected.

21

Search Console and Analytics

First-party search and behavioral data connection state.

Modeled acquisition channels

ChannelEstimated share
direct26.3%
search44.6%
search organic31.1%
search paid13.5%
social8.0%
social organic8.0%
social paid0.0%
referrals11.9%
paid referrals8.0%
mail0.0%
gen ai1.3%
affiliate0.0%
other0.0%

First-party connections

IntegrationStatusData available after connection
Google Search ConsoleNot connectedClicks, impressions, CTR, average position, queries, pages, countries, and devices
Google AnalyticsNot connectedOrganic sessions, engagement, landing pages, conversions, geography, and devices

No first-party Search Console or Analytics numbers are shown until the site owner authorizes the corresponding Google account. Scout traffic model values above are modeled estimates, not private analytics totals.

Limits and method

22

Limitations and open questions

Known constraints and missing evidence from the collected pages.

The collected pages do not mention specific limitations or gaps. However, the pricing page notes that the free plan does not allow commercial use S31. Additionally, the pricing FAQ states that unused credits do not roll over to the next month S39. The site does not name competitors directly, but it publishes comparisons against ElevenLabs and other TTS providers S52.

23

Sources, method, and limitations

Evidence inventory, collection method, provider use, and constraints.

This report is based on 15 collected pages from gradium.ai, including the homepage, pricing, pricing FAQ, and several content articles. The evidence catalog includes 180 source tokens, with 15 pages collected via static and browser rendering rendering. External proof search was enabled but only verified the existence of linked results, not their content. Pricing data was taken from the authoritative pricing page and FAQ. Limitations include that some claims are from marketing content and may not reflect current product state.

Paid against free, in your inbox

One job at a time: the paid tools worth paying for, the free ones you can run yourself, and who should pick which.

No tracking pixels. One click to leave. We use your address to send Stack Radar and nothing else. Privacy.