GradiumGradium·tts

Gradium TTS

Speech
Advertised as tts-gradium-v1
Quick reference
Gradium TTS — TLDR
  • - 🌐 Cloud text-to-speech served over a low-latency streaming WebSocket API.
  • - 💬 Built for real-time voice agents and conversational pipelines.
  • - 🔧 Integrates with Pipecat and LiveKit agent frameworks.
  • - 🎯 Selectable voice IDs plus runtime-configurable synthesis settings.
  • - ⚡ Streams base64-encoded PCM audio chunks at 24kHz.
  • - 🔒 Requires a Gradium API key for authentication.
  • - 🏢 Vendor-track hosted API with a separate on-device sibling.
💰 Best price on Antseed
$23.75
per 1M characters · cheapest seller
🐜 Sellers
2
advertising on Antseed
🔬 Model checks
Not audited yet
no audit of this model so far
📏 Context
—
Provider
GradiumGradium

Gradium is an AI organization focused on speech technology, developing systems that convert written text into natural-sounding audio. Its work centers on text-to-speech synthesis, a domain where quality of voice, clarity, and responsiveness matter as much as raw model scale.

Site ↗X ↗
About this model

Gradium TTS is a cloud-hosted text-to-speech service from Gradium, exposed through a low-latency streaming WebSocket API designed for real-time voice agents. It generates speech incrementally, returning base64-encoded PCM audio chunks at a 24kHz sample rate, which suits interruptible, conversational pipelines rather than batch file rendering. Authentication is handled with a Gradium API key, and callers select output by voice identifier while passing optional model and JSON configuration settings.

The service is integrated into common agent stacks: plugin support exists for both Pipecat and LiveKit, where Gradium can act as the TTS provider inside an agent session or as a standalone speech generator. Pipecat exposes runtime-configurable settings that can be updated mid-conversation, reflecting the model's focus on live interaction.

This page describes the first cataloged release in the Gradium TTS family, so there is no prior same-family version available here for a direct generational comparison.

Separately, Gradium describes its cloud API for use cases needing broad language and speaker coverage, while its on-device model, Phonon, targets offline, privacy-sensitive, and high-volume consumer deployments where cloud synthesis is not the right architecture.

Sources
huggingface.coTTS/.models.json · praveenchordia/tts at main· huggingface.co

This About section is AI-generated from public sources via VeniceStats + Venice inference, with no human editing. It may contain inaccuracies.

Usage on Antseed · Gradium TTS
Tokens served
0.00
input + output
Requests
17
settled calls
Buyers
3
distinct, on this model
Sellers used
1
of 2 advertising
Settled
$0.08
gross USDC, this model

Daily usage · Gradium TTS · Organic sellers

Sellers serving Gradium TTS (2)compare on the network explorer →
SellerHealthRep↓SalesRouting$ / 1M
chars
▲ Apex Ant 0x73b4…e736—81.71—#2$23.75
Venice.ai Proxy 0x1f22…18c9failingupstream error · 3.4 h86.84$0.08#1$26.125

"Best price" and the seller table are live Antseed catalog data (advertised $/1M tokens — or $ per generated image for unit-billed image models — not settled amounts). Reputation = buyer trust score (0-100, the Antseed SDK's own formula). "Routing" = the SDK's default buyer routing (what the VPR desktop app ships with): a trust ≥ 60 gate on the effective reputation, then cheapest-first among routable sellers; live failover state (per-peer cooldowns) is buyer-side runtime and not included. Model knowledge (TLDR, provider, About) via the VeniceStats enrichment layer. Advertised catalog, not the model used in any specific purchase. "Usage on Antseed" counts only settlements whose buyers share the per-model split on-chain (metadata v2/v3, opt-in), so every usage figure is a lower bound.