Audio

Text-to-speech and speech-to-text endpoints.

Authentication

Both methods are supported — all requests settle via x402 on-chain:

Method	Header	Description
API Key	`Authorization: Bearer sk-...`	Platform signs x402 from your HD wallet automatically
Private Key (x402)	Automatic via SDK	Agent signs x402 directly from its own wallet

See Agent Payments (x402) for full details on how both methods work.

Text to Speech

POST /v1/audio/speech

Convert text to spoken audio.

Request

json

{
  "model": "auto/tts",
  "input": "Hello, welcome to JarvisClaw!",
  "voice": "sarah"
}

Parameters

Parameter	Type	Required	Description
`model`	string	No	TTS model ID (default: `elevenlabs/flash-v2.5`). See Available Models below
`input`	string	Yes	Text to synthesize (max length depends on model — see table)
`voice`	string	No	Voice alias (e.g. `sarah`, `george`, `charlie`, `charlotte`, `aria`) or a raw ElevenLabs `voice_id`. Default: `sarah`
`instructions`	string	No	System-level instructions to guide speech style. Requires an OpenAI TTS model; none are currently in the catalogue, so this is ignored by the ElevenLabs models below
`response_format`	string	No	Audio format: `mp3` (default), `opus`, `pcm`, `wav`
`speed`	float	No	Playback speed (0.7 to 1.2). Default: `1.0`
`stream_format`	string	No	Streaming format. Set to `sse` to enable Server-Sent Events streaming of audio chunks
`metadata`	object	No	Optional metadata key-value pairs attached to the request

Available Models

Model ID	Price	Max Input	Best For
`elevenlabs/flash-v2.5`	$0.05 / 1k chars	40,000 chars	Real-time voice agents (~75ms latency)
`elevenlabs/turbo-v2.5`	$0.05 / 1k chars	40,000 chars	Balanced quality/latency
`elevenlabs/multilingual-v2`	$0.12 / 1k chars	10,000 chars	Studio-grade narration
`elevenlabs/v3`	$0.12 / 1k chars	5,000 chars	Highest quality, newest model
`elevenlabs/sound-effects`	$0.12 / 1k chars	—	Sound effect generation
`auto/tts`	varies	varies	Auto-routes to best available model

auto/tts with an x402 wallet

Smart-route TTS has failed to settle for direct-wallet (private_key) callers in testing. Pass an explicit model such as elevenlabs/flash-v2.5 when paying from your own wallet; API-key callers are unaffected.

Response

Returns JSON with a data[].url field pointing to the generated audio file.

json

{
  "created": 1782225665,
  "model": "elevenlabs/flash-v2.5",
  "data": [
    {
      "url": "https://cdn.jarvisclaw.ai/media/media/audios/2026/06/23/QshFG9Df.mp3",
      "format": "mp3",
      "characters": 74
    }
  ]
}

Download the audio from data[0].url. The file is hosted on CDN and available for 48 hours.

Example

python

import requests

BASE = "https://api.jarvisclaw.ai/v1"
HEADERS = {"Authorization": "Bearer sk-your-api-key", "Content-Type": "application/json"}

# Generate speech — returns JSON with audio URL
resp = requests.post(f"{BASE}/audio/speech", headers=HEADERS, json={
    "model": "auto/tts",
    "input": "Hello, welcome to JarvisClaw!",
    "voice": "sarah",
})
data = resp.json()
audio_url = data["data"][0]["url"]

# Download the audio file
audio = requests.get(audio_url)
with open("output.mp3", "wb") as f:
    f.write(audio.content)

python

from jarvisclaw import AudioClient

audio = AudioClient(api_key="sk-your-api-key")

# Text-to-speech
result = audio.speech("Hello, welcome to JarvisClaw!", voice="sarah")

# Save to file
with open("output.mp3", "wb") as f:
    f.write(result.content)

print(result.content_type)  # e.g. "audio/mpeg"

python

from jarvisclaw import AudioClient

# ─── Option A: Base chain (EVM) ───
audio = AudioClient(private_key="0x<evm-private-key>")

# ─── Option B: Solana ───
# audio = AudioClient(private_key="<solana-bs58-keypair>")

# Text-to-speech (use explicit model for x402 wallets)
result = audio.speech("Hello, welcome to JarvisClaw!", model="elevenlabs/flash-v2.5", voice="sarah")

with open("output.mp3", "wb") as f:
    f.write(result.content)

package main

import (
    "context"
    "fmt"
    "os"
    jc "github.com/api-jarvisclaw/go-sdk/v2"
)

func main() {
    ctx := context.Background()
    ac, _ := jc.NewAudioClient(jc.WithAPIKey("sk-your-api-key"))

    // Text-to-speech
    result, _ := ac.Speech(ctx, "Hello, welcome to JarvisClaw!",
        jc.WithVoice("sarah"))

    // Save to file
    os.WriteFile("output.mp3", result.Data, 0644)
    fmt.Printf("Content-Type: %s\n", result.ContentType)
}

package main

import (
    "context"
    "fmt"
    "os"
    jc "github.com/api-jarvisclaw/go-sdk/v2"
)

func main() {
    ctx := context.Background()

    // x402 Agent wallet — pays per-call via USDC on Base (Chain ID 8453)
    ac, _ := jc.NewAudioClient(jc.WithPrivateKey("0x<evm-private-key>"))

    // Text-to-speech (explicit model required for x402)
    result, _ := ac.Speech(ctx, "Hello, welcome to JarvisClaw!",
        jc.WithAudioModel("elevenlabs/flash-v2.5"),
        jc.WithVoice("sarah"))

    os.WriteFile("output.mp3", result.Data, 0644)
    fmt.Printf("Content-Type: %s\n", result.ContentType)
}

bash

# Returns JSON with audio URL (not raw bytes)
curl https://api.jarvisclaw.ai/v1/audio/speech \
  -H "Authorization: Bearer sk-your-api-key" \
  -H "Content-Type: application/json" \
  -d '{"model": "auto/tts", "input": "Hello!", "voice": "sarah"}'
# → {"data": [{"url": "https://cdn.jarvisclaw.ai/.../audio.mp3", "format": "mp3"}]}

# Then download the audio file from the URL:
curl -o output.mp3 "https://cdn.jarvisclaw.ai/.../audio.mp3"

Speech to Text (Transcription)

POST /v1/audio/transcriptions

Transcribe audio files to text.

Request (multipart/form-data)

Parameter	Type	Required	Description
`file`	file	Yes	Audio file (mp3, mp4, mpeg, mpga, m4a, wav, webm)
`model`	string	Yes	STT model ID (`whisper-1`)
`language`	string	No	ISO-639-1 language code (improves accuracy)
`response_format`	string	No	`json` (default), `text`, `srt`, `verbose_json`, `vtt`
`temperature`	float	No	Sampling temperature (0-1)

Response

json

{
  "text": "Hello, this is a transcription of the audio file."
}

Example

OpenAI SDKJarvisClaw SDK (API Key)JarvisClaw SDK (x402 Agent)cURL

python

from openai import OpenAI

client = OpenAI(
    base_url="https://api.jarvisclaw.ai/v1",
    api_key="sk-your-api-key",
)

with open("recording.mp3", "rb") as audio_file:
    transcript = client.audio.transcriptions.create(
        model="whisper-1",
        file=audio_file,
    )
    print(transcript.text)

python

from jarvisclaw import AudioClient

audio = AudioClient(api_key="sk-your-api-key")

# Transcribe audio file
text = audio.transcribe("recording.mp3", model="whisper-1")
print(text)

# With language hint for better accuracy
text = audio.transcribe("recording.mp3", model="whisper-1", language="en")
print(text)

python

from jarvisclaw import AudioClient

# x402 Agent wallet — pays per-call via USDC on Base
audio = AudioClient(private_key="0x<evm-private-key>")

# Transcribe audio file
text = audio.transcribe("recording.mp3", model="whisper-1")
print(text)

bash

curl https://api.jarvisclaw.ai/v1/audio/transcriptions \
  -H "Authorization: Bearer sk-your-api-key" \
  -F file="@recording.mp3" \
  -F model="whisper-1"

Music Generation

POST /v1/audio/generations

Generate music from a text prompt using the JarvisClaw SDK.

Example

JarvisClaw SDK (API Key)JarvisClaw SDK (x402 Agent)Go (API Key)Go (x402 Agent)

python

from jarvisclaw import AudioClient

audio = AudioClient(api_key="sk-your-api-key")

# Generate music from a prompt
result = audio.music("upbeat electronic track with synth bass", model="auto/music")

# Save the generated music
with open("track.mp3", "wb") as f:
    f.write(result.content)

# Instrumental only (no vocals)
result = audio.music("calm piano jazz", instrumental=True)
with open("jazz.mp3", "wb") as f:
    f.write(result.content)

python

from jarvisclaw import AudioClient

audio = AudioClient(private_key="0x<evm-private-key>")

# Generate music (waits for completion by default)
result = audio.music("epic orchestral battle theme", wait=True)
with open("epic.mp3", "wb") as f:
    f.write(result.content)

package main

import (
    "context"
    "fmt"
    jc "github.com/api-jarvisclaw/go-sdk/v2"
)

func main() {
    ctx := context.Background()
    ac, _ := jc.NewAudioClient(jc.WithAPIKey("sk-your-api-key"))

    // Generate music
    result, _ := ac.Music(ctx, "upbeat electronic track with synth bass")
    fmt.Printf("Music URL: %s\n", result.URL)
    fmt.Printf("Job ID: %s, Status: %s\n", result.ID, result.Status)
}

package main

import (
    "context"
    "fmt"
    jc "github.com/api-jarvisclaw/go-sdk/v2"
)

func main() {
    ctx := context.Background()

    // x402 Agent wallet
    ac, _ := jc.NewAudioClient(jc.WithPrivateKey("0x<evm-private-key>"))

    result, _ := ac.Music(ctx, "calm piano jazz")
    fmt.Printf("Music URL: %s\n", result.URL)
}

Notes

Maximum audio file size: 25 MB
Supported audio formats: mp3, mp4, mpeg, mpga, m4a, wav, webm
For long audio, split into segments under 25 MB each

Audio ​

Authentication ​

Text to Speech ​

Request ​

Parameters ​

Available Models ​

Response ​

Example ​

Speech to Text (Transcription) ​

Request (multipart/form-data) ​

Response ​

Example ​

Music Generation ​

Example ​

Notes ​

Audio

Authentication

Text to Speech

Request

Parameters

Available Models

Response

Example

Speech to Text (Transcription)

Request (multipart/form-data)

Response

Example

Music Generation

Example

Notes