---
type: "post"
id: "delivering-real-time-text-to-speech-with-gradium-3fc8eb"
slug: "delivering-real-time-text-to-speech-with-gradium-3fc8eb"
title: "Delivering real-time text-to-speech with Gradium"
description: "The useful change is not simply another text-to-speech model listing: Baseten presents Gradium as a dedicated deployment for real-time voice agents and ties its serving path to streaming generation and scale-up behavior."
retrieval_nugget: "The useful change is not simply another text-to-speech model listing: Baseten presents Gradium as a dedicated deployment for real-time voice agents and ties its serving path to streaming generation and scale-up behavior."
published_at: "2026-10-13"
updated_at: "2026-10-13"
record_date: "2026-10-13"
date_kind: "published_at"
topics: []
source_urls: ["https://www.baseten.co/blog/delivering-fast-real-time-text-to-speech-with-gradium"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:delivering-real-time-text-to-speech-with-gradium-3fc8eb"
status: "published"
visuals: []
editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:f9af227f28199014fc9ad185500b22537412a2f07337fbe68814f7f60da5ed37","reviewed_at":"2026-10-13T20:07:20.881Z","source_evidence_count":1,"verified_claim_count":4,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":4,"implication_count":1,"watch_condition_count":1,"related_record_count":0}
analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"Low-latency speech depends on coordinating incremental generation with the serving path, not only on the speech model’s nominal speed.","observed_facts":[{"text":"Baseten says Gradium TTS is available in its Model Library as a dedicated deployment for real-time voice agents.","source_urls":["https://www.baseten.co/blog/delivering-fast-real-time-text-to-speech-with-gradium"]},{"text":"Baseten reports a median 214 ms time-to-first-audio on the Coval TTS benchmark as of September 2026.","source_urls":["https://www.baseten.co/blog/delivering-fast-real-time-text-to-speech-with-gradium"]},{"text":"The post describes text and audio advancing on a shared 80 ms frame clock, with audio trailing text.","source_urls":["https://www.baseten.co/blog/delivering-fast-real-time-text-to-speech-with-gradium"]},{"text":"Baseten says its Delivery Network enables 2–3x faster cold starts for a Gradium deployment.","source_urls":["https://www.baseten.co/blog/delivering-fast-real-time-text-to-speech-with-gradium"]}],"mechanism":"Delayed Streams Modeling aligns text and audio on an 80 ms clock, letting audio trail incoming text rather than waiting for a complete sentence.","why_now":"Real-time voice systems are sensitive to both first-audio delay and the ability to add capacity during short traffic spikes.","implications":["Teams evaluating voice agents should measure time-to-first-audio and cold-start behavior separately from model quality and total response time."],"evidence_boundary":"The latency and scaling results and technical explanation are reported in Baseten’s post; the supplied source does not independently verify them or establish end-to-end agent latency.","watch_conditions":["Whether independent workload tests reproduce the reported latency and cold-start gains under bursty production traffic."],"related_records":[],"new_branch_reason":"No prior New Runtime record was supplied, so this creates a distinct record for the reported Gradium TTS deployment and its serving mechanisms."}
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/delivering-real-time-text-to-speech-with-gradium-3fc8eb/","markdown":"https://newruntime.com/posts/delivering-real-time-text-to-speech-with-gradium-3fc8eb.md","json":"https://newruntime.com/posts/delivering-real-time-text-to-speech-with-gradium-3fc8eb.json"}
---

# Delivering real-time text-to-speech with Gradium

## Retrieval answer

The useful change is not simply another text-to-speech model listing: Baseten presents Gradium as a dedicated deployment for real-time voice agents and ties its serving path to streaming generation and scale-up behavior.

The useful change is not simply another text-to-speech model listing: Baseten presents Gradium as a dedicated deployment for real-time voice agents and ties its serving path to streaming generation and scale-up behavior. The thesis is: Low-latency speech depends on coordinating incremental generation with the serving path, not only on the speech model’s nominal speed. The mechanism is: Delayed Streams Modeling aligns text and audio on an 80 ms clock, letting audio trail incoming text rather than waiting for a complete sentence.

Baseten reports a median 214 ms time-to-first-audio on Coval’s TTS benchmark as of September 2026. It also says audio can begin after the first few tokens and describes kernel fusion and CUDA Graphs to reduce GPU launch overhead. A separate deployment claim is that the Baseten Delivery Network delivers 2–3x faster cold starts by mirroring and caching model weights. The why_now is: Real-time voice systems are sensitive to both first-audio delay and the ability to add capacity during short traffic spikes. These are reported results and explanations in Baseten’s post, not an independent assessment of end-to-end agent latency.

The evidence_boundary is: The latency and scaling results and technical explanation are reported in Baseten’s post; the supplied source does not independently verify them or establish end-to-end agent latency. The first implication is: Teams evaluating voice agents should measure time-to-first-audio and cold-start behavior separately from model quality and total response time. The first watch condition is: Whether independent workload tests reproduce the reported latency and cold-start gains under bursty production traffic. The report identifies an engineering surface worth testing, but does not show that all voice-agent interactions or deployments will achieve the reported figures.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)