Caedral Voice

By Caedral

Caedral Voice 1

Natural English speech through one simple API. Four production voices, WAV output, billed on your Caedral Models balance.

caedral-voice-1 · Text → Speech · $0.03 / 1k chars

Get API key Documentation

Caedral F1English

Voice showcase

Four production voices.

Pre-rendered samples. They do not call the Voice API and do not debit any balance.

Caedral F1
FemaleEnglish
caedral-f1
Caedral F2
FemaleEnglish
caedral-f2
Caedral M1
MaleEnglish
caedral-m1
Caedral M2
MaleEnglish
caedral-m2

Ownership

A Caedral model, not a wrapper label.

caedral-voice-1 is developed and served by Caedral. It pays from plan credits and prepaid balance. The separate catalog id caedral-voice is a third-party convenience product over OpenAI gpt-audio and bills your prepaid balance.

Why Caedral Voice

Speech that fits the rest of your stack.

One Caedral API key — the same account you already use for chat, embeddings, and rerank.
Natural English voices ready for products, not a research demo.
Four production identities: F1, F2, M1, and M2.
Straightforward pricing on the Caedral Models balance — no separate Voice pool.
Official docs, SDKs, and an n8n node for the same endpoint.

Use cases

Where teams use it.

AI assistants
Narration
Applications
Notifications
Accessibility
Voice interfaces
Content workflows
Automation

Pricing

$0.03 / 1k characters.

Usage is measured in characters and paid from your weekly plan credits first, then your prepaid balance. There is no separate Voice quota.

Customer price: $0.03 / 1k chars
Billing unit: Characters
Pays with: Plan credits or prepaid balance

Every plan includes weekly credits, and the Free plan's voices (F1 + F2) include Caedral Voice. Usage past the weekly allowance flows to your prepaid balance.

Capabilities

What V1 supports.

Model ID: caedral-voice-1
Language: English (V1)
Voices: caedral-f1, caedral-f2, caedral-m1, caedral-m2
Output: WAV, 24 kHz
Max input: 2000 characters
Speed: 0.5–2
Endpoint: POST /v1/audio/speech

Developer experience

A short request. A WAV file back.

Authenticate with a Caedral API key. Choose a voice id, send text, receive WAV. Speed 0.5–2.0. Full request and response details live in the docs.

POST /v1/audio/speechbash

1curl https://api.caedral.com/v1/audio/speech \2  -H "Authorization: Bearer $CAEDRAL_API_KEY" \3  -H "Content-Type: application/json" \4  -d '{"model":"caedral-voice-1","voice":"caedral-f1","input":"Welcome to Caedral.","speed":1.0,"response_format":"wav"}' \5  --output welcome.wav

Integrations

API, SDKs, and n8n.

The same public speech endpoint works from REST, official SDKs, and the Caedral n8n node.

API
POST /v1/audio/speech
SDKs
TypeScript, Python, Go, Java, and more where the speech helper exists.
n8n
Official Caedral community node.

Status

Availability

Live catalog status for caedral-voice-1. Internal health URLs are not published here.

Operational

Attribution

Kokoro foundation

Caedral Voice is a Caedral product built on an attributed open-weight Kokoro foundation (Apache-2.0). The base Kokoro weights were not trained by Caedral.