Kokoro 82m API
Run hexgrad/kokoro-82m through FastInfra's API.
Text-to-speech — billed per input character on the same FastInfra API key and base URL (POST /audio/speech).
API access for hexgrad/kokoro-82m
One Base URL and API key for every model. Use the endpoint that matches the capability — not every model uses chat completions.
| Field | Value |
|---|---|
| Model ID | hexgrad/kokoro-82m |
| Capability | Text-to-speech |
| Base URL | https://api.fastinfra.ai/v1 |
| Primary endpoint | POST https://api.fastinfra.ai/v1/audio/speech |
| Auth | Authorization: Bearer YOUR_API_KEY |
| OpenAI Chat SDK | No — use the endpoint above |
| Validate model id | GET https://api.fastinfra.ai/v1/models/hexgrad/kokoro-82m |
| Docs | Integration reference |
Returns audio bytes; not chat completions.
Kokoro 82m pricing
Billed per input character: 1 character = 1 output token, so the output rate below is the price per 1M characters. 50% below the cheapest hosted provider.
| Direction | Price per 1M tokens |
|---|---|
| Input | $0.01 |
| Output | $0.30 |
| Per 1M characters | $0.30 |
1 route available
Requests route to fastinfra by default (lowest cost). FastInfra fails over automatically if a route is unavailable.
| Route label | Model ID |
|---|---|
fastinfra |
hexgrad/kokoro-82m |
Call Kokoro 82m in 30 seconds
POST /audio/speech with this model id and get WAV back — see Text-to-speech in the docs.
Python
from openai import OpenAI
client = OpenAI(api_key="YOUR_API_KEY", base_url="https://api.fastinfra.ai/v1")
with client.audio.speech.with_streaming_response.create(
model="tts-1",
voice="af_heart",
input="Welcome to FastInfra.",
) as response:
response.stream_to_file("speech.wav")curl
curl https://api.fastinfra.ai/v1/audio/speech \
-H "Authorization: Bearer YOUR_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "tts-1",
"voice": "af_heart",
"input": "Welcome to FastInfra."
}' --output speech.wavKokoro 82m — common questions
How much does the hexgrad/kokoro-82m API cost?
On FastInfra, hexgrad/kokoro-82m costs $0.01/1M input, $0.3/1M output tokens. Billing is per token used, with no subscription.
Is hexgrad/kokoro-82m compatible with the OpenAI SDK?
Yes. FastInfra exposes hexgrad/kokoro-82m through an OpenAI-compatible endpoint at https://api.fastinfra.ai/v1 — point any OpenAI SDK at that base URL with a FastInfra API key and keep your existing code.
How does routing work for hexgrad/kokoro-82m?
hexgrad/kokoro-82m is available on 1 route(s): fastinfra. Requests use fastinfra by default (lowest cost). FastInfra fails over automatically if a route is unavailable.
Related models
Compare side by side: Kokoro 82m vs Gemma2:2b · Kokoro 82m vs Mxbai Embed Large · Kokoro 82m vs Nomic Embed Text