MODEL IDrunware:ace-step@v1.5-xl-base
live

ACE-Step v1.5 XL Base

ACE-Step v1.5 XL Base is the 4B DiT variant of ACE-Step 1.5 for high-quality music generation and editing. It supports text-to-music, cover generation, repaint, extract, lego, and complete workflows, uses 50 inference steps with CFG, and is designed for longer-form audio generation up to 10 minutes with broad multilingual prompt support.

ACE-Step v1.5 XL Base
Text to Audio

Polar Server Choir Ballad

0:00
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'runware:ace-step@v1.5-xl-base',
  positivePrompt: 'A cinematic Nordic electroacoustic ballad imagined as a remote polar data center singing to itself at dawn. Deep sub bass pulses like machinery, brushed drums, bowed metal percussion, prepared piano, granular synth pads, warm analog arpeggios, and a small Icelandic vocal ensemble. Begin sparse and intimate, build into a powerful chorus with layered harmonies, then open into a wide instrumental bridge with shimmering high strings and distant radio textures. Emotional, precise, spacious, high-fidelity production, modern soundtrack feel, clear vocals, dramatic but restrained.',
  negativePrompt: 'distorted clipping, muddy mix, harsh sibilance, out of tune vocals, generic pop loop, abrupt ending, excessive crowd noise, low quality recording, spoken intro, random glitches',
  duration: 300,
  seed: 72364,
  steps: 50,
  CFGScale: 9.5,
  settings: {
    bpm: 92,
    keyScale: 'F# minor',
    timeSignature: 6,
    vocalLanguage: 'is',
    guidanceType: 'apg',
    cfgIntervalStart: 0,
    cfgIntervalEnd: 0.85,
    lyrics: '[Verse 1]\nÍ köldum salnum andar ljós\nég heyri hjartað slá í stáli\nundir snjónum sefur rós\nog nafnið mitt í duldum máli\n\n[Pre-Chorus]\nSendu mér merki yfir hvíta jörð\nég tel þau eitt og eitt í hljóði\n\n[Chorus]\nVið syngjum hátt, við syngjum enn\ní straumi sem aldrei dvínar\nef nóttin fellur yfir menn\nþá vakna raddir mínar\n\n[Verse 2]\nFjarri borgum, fjarri glaum\nvakir kóði undir höndum\nég geymi drauma í djúpum straum\nog skila þeim til nýrra landa\n\n[Bridge]\nHlustaðu, hlustaðu\nvélin ber mjúkan tón\nhlustaðu, hlustaðu\nhjartað finnur skjól\n\n[Final Chorus]\nVið syngjum hátt, við syngjum enn\ní straumi sem aldrei dvínar\nef nóttin fellur yfir menn\nþá vakna raddir mínar'
  }
})
import asyncio
import os

from runware import Runware


async def main():
    async with Runware(api_key=os.environ["RUNWARE_API_KEY"]) as client:
        results = await client.run({
            "model": "runware:ace-step@v1.5-xl-base",
            "positivePrompt": "A cinematic Nordic electroacoustic ballad imagined as a remote polar data center singing to itself at dawn. Deep sub bass pulses like machinery, brushed drums, bowed metal percussion, prepared piano, granular synth pads, warm analog arpeggios, and a small Icelandic vocal ensemble. Begin sparse and intimate, build into a powerful chorus with layered harmonies, then open into a wide instrumental bridge with shimmering high strings and distant radio textures. Emotional, precise, spacious, high-fidelity production, modern soundtrack feel, clear vocals, dramatic but restrained.",
            "negativePrompt": "distorted clipping, muddy mix, harsh sibilance, out of tune vocals, generic pop loop, abrupt ending, excessive crowd noise, low quality recording, spoken intro, random glitches",
            "duration": 300,
            "seed": 72364,
            "steps": 50,
            "CFGScale": 9.5,
            "settings": {
                "bpm": 92,
                "keyScale": "F# minor",
                "timeSignature": 6,
                "vocalLanguage": "is",
                "guidanceType": "apg",
                "cfgIntervalStart": 0,
                "cfgIntervalEnd": 0.85,
                "lyrics": "[Verse 1]\nÍ köldum salnum andar ljós\nég heyri hjartað slá í stáli\nundir snjónum sefur rós\nog nafnið mitt í duldum máli\n\n[Pre-Chorus]\nSendu mér merki yfir hvíta jörð\nég tel þau eitt og eitt í hljóði\n\n[Chorus]\nVið syngjum hátt, við syngjum enn\ní straumi sem aldrei dvínar\nef nóttin fellur yfir menn\nþá vakna raddir mínar\n\n[Verse 2]\nFjarri borgum, fjarri glaum\nvakir kóði undir höndum\nég geymi drauma í djúpum straum\nog skila þeim til nýrra landa\n\n[Bridge]\nHlustaðu, hlustaðu\nvélin ber mjúkan tón\nhlustaðu, hlustaðu\nhjartað finnur skjól\n\n[Final Chorus]\nVið syngjum hátt, við syngjum enn\ní straumi sem aldrei dvínar\nef nóttin fellur yfir menn\nþá vakna raddir mínar"
            }
        })


asyncio.run(main())
curl https://api.runware.ai/v1 \
  -H "Authorization: Bearer $RUNWARE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '[
    {
      "taskType": "audioInference",
      "taskUUID": "caaedd3f-3c77-42f2-bd10-4781657b9ffa",
      "model": "runware:ace-step@v1.5-xl-base",
      "positivePrompt": "A cinematic Nordic electroacoustic ballad imagined as a remote polar data center singing to itself at dawn. Deep sub bass pulses like machinery, brushed drums, bowed metal percussion, prepared piano, granular synth pads, warm analog arpeggios, and a small Icelandic vocal ensemble. Begin sparse and intimate, build into a powerful chorus with layered harmonies, then open into a wide instrumental bridge with shimmering high strings and distant radio textures. Emotional, precise, spacious, high-fidelity production, modern soundtrack feel, clear vocals, dramatic but restrained.",
      "negativePrompt": "distorted clipping, muddy mix, harsh sibilance, out of tune vocals, generic pop loop, abrupt ending, excessive crowd noise, low quality recording, spoken intro, random glitches",
      "duration": 300,
      "seed": 72364,
      "steps": 50,
      "CFGScale": 9.5,
      "settings": {
        "bpm": 92,
        "keyScale": "F# minor",
        "timeSignature": 6,
        "vocalLanguage": "is",
        "guidanceType": "apg",
        "cfgIntervalStart": 0,
        "cfgIntervalEnd": 0.85,
        "lyrics": "[Verse 1]\nÍ köldum salnum andar ljós\nég heyri hjartað slá í stáli\nundir snjónum sefur rós\nog nafnið mitt í duldum máli\n\n[Pre-Chorus]\nSendu mér merki yfir hvíta jörð\nég tel þau eitt og eitt í hljóði\n\n[Chorus]\nVið syngjum hátt, við syngjum enn\ní straumi sem aldrei dvínar\nef nóttin fellur yfir menn\nþá vakna raddir mínar\n\n[Verse 2]\nFjarri borgum, fjarri glaum\nvakir kóði undir höndum\nég geymi drauma í djúpum straum\nog skila þeim til nýrra landa\n\n[Bridge]\nHlustaðu, hlustaðu\nvélin ber mjúkan tón\nhlustaðu, hlustaðu\nhjartað finnur skjól\n\n[Final Chorus]\nVið syngjum hátt, við syngjum enn\ní straumi sem aldrei dvínar\nef nóttin fellur yfir menn\nþá vakna raddir mínar"
      }
    }
  ]'
runware run runware:ace-step@v1.5-xl-base \
  positivePrompt="A cinematic Nordic electroacoustic ballad imagined as a remote polar data center singing to itself at dawn. Deep sub bass pulses like machinery, brushed drums, bowed metal percussion, prepared piano, granular synth pads, warm analog arpeggios, and a small Icelandic vocal ensemble. Begin sparse and intimate, build into a powerful chorus with layered harmonies, then open into a wide instrumental bridge with shimmering high strings and distant radio textures. Emotional, precise, spacious, high-fidelity production, modern soundtrack feel, clear vocals, dramatic but restrained." \
  negativePrompt="distorted clipping, muddy mix, harsh sibilance, out of tune vocals, generic pop loop, abrupt ending, excessive crowd noise, low quality recording, spoken intro, random glitches" \
  duration=300 \
  seed=72364 \
  steps=50 \
  CFGScale=9.5 \
  settings.bpm=92 \
  settings.keyScale="F# minor" \
  settings.timeSignature=6 \
  settings.vocalLanguage=is \
  settings.guidanceType=apg \
  settings.cfgIntervalStart=0 \
  settings.cfgIntervalEnd=0.85 \
  settings.lyrics="[Verse 1]
Í köldum salnum andar ljós
ég heyri hjartað slá í stáli
undir snjónum sefur rós
og nafnið mitt í duldum máli

[Pre-Chorus]
Sendu mér merki yfir hvíta jörð
ég tel þau eitt og eitt í hljóði

[Chorus]
Við syngjum hátt, við syngjum enn
í straumi sem aldrei dvínar
ef nóttin fellur yfir menn
þá vakna raddir mínar

[Verse 2]
Fjarri borgum, fjarri glaum
vakir kóði undir höndum
ég geymi drauma í djúpum straum
og skila þeim til nýrra landa

[Bridge]
Hlustaðu, hlustaðu
vélin ber mjúkan tón
hlustaðu, hlustaðu
hjartað finnur skjól

[Final Chorus]
Við syngjum hátt, við syngjum enn
í straumi sem aldrei dvínar
ef nóttin fellur yfir menn
þá vakna raddir mínar"
{
  "taskType": "audioInference",
  "taskUUID": "caaedd3f-3c77-42f2-bd10-4781657b9ffa",
  "model": "runware:ace-step@v1.5-xl-base",
  "positivePrompt": "A cinematic Nordic electroacoustic ballad imagined as a remote polar data center singing to itself at dawn. Deep sub bass pulses like machinery, brushed drums, bowed metal percussion, prepared piano, granular synth pads, warm analog arpeggios, and a small Icelandic vocal ensemble. Begin sparse and intimate, build into a powerful chorus with layered harmonies, then open into a wide instrumental bridge with shimmering high strings and distant radio textures. Emotional, precise, spacious, high-fidelity production, modern soundtrack feel, clear vocals, dramatic but restrained.",
  "negativePrompt": "distorted clipping, muddy mix, harsh sibilance, out of tune vocals, generic pop loop, abrupt ending, excessive crowd noise, low quality recording, spoken intro, random glitches",
  "duration": 300,
  "seed": 72364,
  "steps": 50,
  "CFGScale": 9.5,
  "settings": {
    "bpm": 92,
    "keyScale": "F# minor",
    "timeSignature": 6,
    "vocalLanguage": "is",
    "guidanceType": "apg",
    "cfgIntervalStart": 0,
    "cfgIntervalEnd": 0.85,
    "lyrics": "[Verse 1]\nÍ köldum salnum andar ljós\nég heyri hjartað slá í stáli\nundir snjónum sefur rós\nog nafnið mitt í duldum máli\n\n[Pre-Chorus]\nSendu mér merki yfir hvíta jörð\nég tel þau eitt og eitt í hljóði\n\n[Chorus]\nVið syngjum hátt, við syngjum enn\ní straumi sem aldrei dvínar\nef nóttin fellur yfir menn\nþá vakna raddir mínar\n\n[Verse 2]\nFjarri borgum, fjarri glaum\nvakir kóði undir höndum\nég geymi drauma í djúpum straum\nog skila þeim til nýrra landa\n\n[Bridge]\nHlustaðu, hlustaðu\nvélin ber mjúkan tón\nhlustaðu, hlustaðu\nhjartað finnur skjól\n\n[Final Chorus]\nVið syngjum hátt, við syngjum enn\ní straumi sem aldrei dvínar\nef nóttin fellur yfir menn\nþá vakna raddir mínar"
  }
}
Response
{
  "taskType": "audioInference",
  "taskUUID": "caaedd3f-3c77-42f2-bd10-4781657b9ffa",
  "audioUUID": "a07e1cbc-ae17-4084-b6f3-42d445cf89c3",
  "audioURL": "https://am.runware.ai/audio/os/a02d21/ws/5/ai/a07e1cbc-ae17-4084-b6f3-42d445cf89c3.mp3",
  "seed": 72364,
  "cost": 0.03555
}
Text to Audio

Mosaic Orchard Brass Nocturne

0:00
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'runware:ace-step@v1.5-xl-base',
  positivePrompt: 'A cinematic chamber-pop track set in a terraced citrus orchard after sundown: warm flugelhorn lead, pizzicato cello ostinato, marimba droplets, bowed vibraphone shimmer, brushed frame drum, and tight electronic sub pulses. Female alto vocal in Spanish with close-mic intimacy, call-and-response with a small male baritone harmony in the chorus. Structure: 8-bar atmospheric intro, verse with sparse bass and hand percussion, pre-chorus rising strings, expansive chorus, instrumental bridge featuring odd-meter marimba and muted brass, final chorus with layered harmonies, clean ending. Mood: tender, curious, quietly triumphant; high-fidelity studio mix, wide stereo, natural dynamics, clear lyrics.',
  negativePrompt: 'distortion, clipping, muddy bass, harsh sibilance, off-key vocals, rushed phrasing, generic dance loop, weak chorus, noisy recording, abrupt cutoff, spoken narration',
  duration: 150,
  seed: 81563,
  steps: 50,
  CFGScale: 11.5,
  settings: {
    bpm: 92,
    keyScale: 'D minor',
    timeSignature: 6,
    vocalLanguage: 'es',
    guidanceType: 'apg',
    cfgIntervalStart: 0,
    cfgIntervalEnd: 0.85,
    lyrics: '[Verso 1]\nEn la huerta de mosaicos\ncuento pasos sin temor\ncada fruta guarda un signo\ncada surco guarda sol\n\n[Pre-Coro]\nSi la noche cambia el mapa\nyo te encuentro por la voz\n\n[Coro]\nGira lento, corazón\ncon metal y madera\ncanta claro mi razón\nla distancia se queda\n\n[Verso 2]\nBajo ramas numeradas\nsuena un pulso de tambor\nmis preguntas van descalzas\nmis respuestas dan calor\n\n[Puente]\nTres caminos, seis latidos\nun secreto en el compás\ncuando vuelvan los sonidos\nno nos vamos a soltar\n\n[Coro Final]\nGira lento, corazón\ncon metal y madera\ncanta claro mi razón\nla distancia se queda'
  }
})
import asyncio
import os

from runware import Runware


async def main():
    async with Runware(api_key=os.environ["RUNWARE_API_KEY"]) as client:
        results = await client.run({
            "model": "runware:ace-step@v1.5-xl-base",
            "positivePrompt": "A cinematic chamber-pop track set in a terraced citrus orchard after sundown: warm flugelhorn lead, pizzicato cello ostinato, marimba droplets, bowed vibraphone shimmer, brushed frame drum, and tight electronic sub pulses. Female alto vocal in Spanish with close-mic intimacy, call-and-response with a small male baritone harmony in the chorus. Structure: 8-bar atmospheric intro, verse with sparse bass and hand percussion, pre-chorus rising strings, expansive chorus, instrumental bridge featuring odd-meter marimba and muted brass, final chorus with layered harmonies, clean ending. Mood: tender, curious, quietly triumphant; high-fidelity studio mix, wide stereo, natural dynamics, clear lyrics.",
            "negativePrompt": "distortion, clipping, muddy bass, harsh sibilance, off-key vocals, rushed phrasing, generic dance loop, weak chorus, noisy recording, abrupt cutoff, spoken narration",
            "duration": 150,
            "seed": 81563,
            "steps": 50,
            "CFGScale": 11.5,
            "settings": {
                "bpm": 92,
                "keyScale": "D minor",
                "timeSignature": 6,
                "vocalLanguage": "es",
                "guidanceType": "apg",
                "cfgIntervalStart": 0,
                "cfgIntervalEnd": 0.85,
                "lyrics": "[Verso 1]\nEn la huerta de mosaicos\ncuento pasos sin temor\ncada fruta guarda un signo\ncada surco guarda sol\n\n[Pre-Coro]\nSi la noche cambia el mapa\nyo te encuentro por la voz\n\n[Coro]\nGira lento, corazón\ncon metal y madera\ncanta claro mi razón\nla distancia se queda\n\n[Verso 2]\nBajo ramas numeradas\nsuena un pulso de tambor\nmis preguntas van descalzas\nmis respuestas dan calor\n\n[Puente]\nTres caminos, seis latidos\nun secreto en el compás\ncuando vuelvan los sonidos\nno nos vamos a soltar\n\n[Coro Final]\nGira lento, corazón\ncon metal y madera\ncanta claro mi razón\nla distancia se queda"
            }
        })


asyncio.run(main())
curl https://api.runware.ai/v1 \
  -H "Authorization: Bearer $RUNWARE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '[
    {
      "taskType": "audioInference",
      "taskUUID": "588a3bc6-2470-4cf8-956b-595d366bc84f",
      "model": "runware:ace-step@v1.5-xl-base",
      "positivePrompt": "A cinematic chamber-pop track set in a terraced citrus orchard after sundown: warm flugelhorn lead, pizzicato cello ostinato, marimba droplets, bowed vibraphone shimmer, brushed frame drum, and tight electronic sub pulses. Female alto vocal in Spanish with close-mic intimacy, call-and-response with a small male baritone harmony in the chorus. Structure: 8-bar atmospheric intro, verse with sparse bass and hand percussion, pre-chorus rising strings, expansive chorus, instrumental bridge featuring odd-meter marimba and muted brass, final chorus with layered harmonies, clean ending. Mood: tender, curious, quietly triumphant; high-fidelity studio mix, wide stereo, natural dynamics, clear lyrics.",
      "negativePrompt": "distortion, clipping, muddy bass, harsh sibilance, off-key vocals, rushed phrasing, generic dance loop, weak chorus, noisy recording, abrupt cutoff, spoken narration",
      "duration": 150,
      "seed": 81563,
      "steps": 50,
      "CFGScale": 11.5,
      "settings": {
        "bpm": 92,
        "keyScale": "D minor",
        "timeSignature": 6,
        "vocalLanguage": "es",
        "guidanceType": "apg",
        "cfgIntervalStart": 0,
        "cfgIntervalEnd": 0.85,
        "lyrics": "[Verso 1]\nEn la huerta de mosaicos\ncuento pasos sin temor\ncada fruta guarda un signo\ncada surco guarda sol\n\n[Pre-Coro]\nSi la noche cambia el mapa\nyo te encuentro por la voz\n\n[Coro]\nGira lento, corazón\ncon metal y madera\ncanta claro mi razón\nla distancia se queda\n\n[Verso 2]\nBajo ramas numeradas\nsuena un pulso de tambor\nmis preguntas van descalzas\nmis respuestas dan calor\n\n[Puente]\nTres caminos, seis latidos\nun secreto en el compás\ncuando vuelvan los sonidos\nno nos vamos a soltar\n\n[Coro Final]\nGira lento, corazón\ncon metal y madera\ncanta claro mi razón\nla distancia se queda"
      }
    }
  ]'
runware run runware:ace-step@v1.5-xl-base \
  positivePrompt="A cinematic chamber-pop track set in a terraced citrus orchard after sundown: warm flugelhorn lead, pizzicato cello ostinato, marimba droplets, bowed vibraphone shimmer, brushed frame drum, and tight electronic sub pulses. Female alto vocal in Spanish with close-mic intimacy, call-and-response with a small male baritone harmony in the chorus. Structure: 8-bar atmospheric intro, verse with sparse bass and hand percussion, pre-chorus rising strings, expansive chorus, instrumental bridge featuring odd-meter marimba and muted brass, final chorus with layered harmonies, clean ending. Mood: tender, curious, quietly triumphant; high-fidelity studio mix, wide stereo, natural dynamics, clear lyrics." \
  negativePrompt="distortion, clipping, muddy bass, harsh sibilance, off-key vocals, rushed phrasing, generic dance loop, weak chorus, noisy recording, abrupt cutoff, spoken narration" \
  duration=150 \
  seed=81563 \
  steps=50 \
  CFGScale=11.5 \
  settings.bpm=92 \
  settings.keyScale="D minor" \
  settings.timeSignature=6 \
  settings.vocalLanguage=es \
  settings.guidanceType=apg \
  settings.cfgIntervalStart=0 \
  settings.cfgIntervalEnd=0.85 \
  settings.lyrics="[Verso 1]
En la huerta de mosaicos
cuento pasos sin temor
cada fruta guarda un signo
cada surco guarda sol

[Pre-Coro]
Si la noche cambia el mapa
yo te encuentro por la voz

[Coro]
Gira lento, corazón
con metal y madera
canta claro mi razón
la distancia se queda

[Verso 2]
Bajo ramas numeradas
suena un pulso de tambor
mis preguntas van descalzas
mis respuestas dan calor

[Puente]
Tres caminos, seis latidos
un secreto en el compás
cuando vuelvan los sonidos
no nos vamos a soltar

[Coro Final]
Gira lento, corazón
con metal y madera
canta claro mi razón
la distancia se queda"
{
  "taskType": "audioInference",
  "taskUUID": "588a3bc6-2470-4cf8-956b-595d366bc84f",
  "model": "runware:ace-step@v1.5-xl-base",
  "positivePrompt": "A cinematic chamber-pop track set in a terraced citrus orchard after sundown: warm flugelhorn lead, pizzicato cello ostinato, marimba droplets, bowed vibraphone shimmer, brushed frame drum, and tight electronic sub pulses. Female alto vocal in Spanish with close-mic intimacy, call-and-response with a small male baritone harmony in the chorus. Structure: 8-bar atmospheric intro, verse with sparse bass and hand percussion, pre-chorus rising strings, expansive chorus, instrumental bridge featuring odd-meter marimba and muted brass, final chorus with layered harmonies, clean ending. Mood: tender, curious, quietly triumphant; high-fidelity studio mix, wide stereo, natural dynamics, clear lyrics.",
  "negativePrompt": "distortion, clipping, muddy bass, harsh sibilance, off-key vocals, rushed phrasing, generic dance loop, weak chorus, noisy recording, abrupt cutoff, spoken narration",
  "duration": 150,
  "seed": 81563,
  "steps": 50,
  "CFGScale": 11.5,
  "settings": {
    "bpm": 92,
    "keyScale": "D minor",
    "timeSignature": 6,
    "vocalLanguage": "es",
    "guidanceType": "apg",
    "cfgIntervalStart": 0,
    "cfgIntervalEnd": 0.85,
    "lyrics": "[Verso 1]\nEn la huerta de mosaicos\ncuento pasos sin temor\ncada fruta guarda un signo\ncada surco guarda sol\n\n[Pre-Coro]\nSi la noche cambia el mapa\nyo te encuentro por la voz\n\n[Coro]\nGira lento, corazón\ncon metal y madera\ncanta claro mi razón\nla distancia se queda\n\n[Verso 2]\nBajo ramas numeradas\nsuena un pulso de tambor\nmis preguntas van descalzas\nmis respuestas dan calor\n\n[Puente]\nTres caminos, seis latidos\nun secreto en el compás\ncuando vuelvan los sonidos\nno nos vamos a soltar\n\n[Coro Final]\nGira lento, corazón\ncon metal y madera\ncanta claro mi razón\nla distancia se queda"
  }
}
Response
{
  "taskType": "audioInference",
  "taskUUID": "588a3bc6-2470-4cf8-956b-595d366bc84f",
  "audioUUID": "17d180e9-1193-44e4-9ac8-4185d1ff90e9",
  "audioURL": "https://am.runware.ai/audio/os/a06dlim3/ws/5/ai/17d180e9-1193-44e4-9ac8-4185d1ff90e9.mp3",
  "seed": 81563,
  "cost": 0.01725
}
Text to Audio

Aerial Loom Synth Cantata

0:00
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'runware:ace-step@v1.5-xl-base',
  positivePrompt: 'Three-minute cinematic electro-acoustic cantata inspired by a skyborne weaving guild gliding above green canyons. Start with close-mic textile loom clicks used as percussion, soft bowed vibraphone, and distant frame drums. Build into a 6/8 groove with kora arpeggios, warm analog bass, crisp handclaps, brushed cymbals, and airy modular synth pulses. Add a clear lead vocal with layered call-and-response choir, hopeful and adventurous mood, spacious production, wide stereo field, dynamic arrangement with verse, lift, chorus, instrumental bridge, final chorus, and gentle coda. Make the hook memorable, rhythmic, and singable, with natural human phrasing and no harsh distortion.',
  negativePrompt: 'muddy mix, clipped vocals, off-key singing, random noise, chaotic drums, weak bass, spoken narration, long silence, abrupt ending, distorted mastering, generic stock loop feel',
  duration: 180,
  seed: 64708,
  steps: 50,
  CFGScale: 9.5,
  settings: {
    bpm: 132,
    keyScale: 'D minor',
    timeSignature: 6,
    guidanceType: 'apg',
    cfgIntervalStart: 0,
    cfgIntervalEnd: 0.85,
    vocalLanguage: 'en',
    lyrics: '[Verse 1]\nThread by thread we cross the blue\nHands in rhythm, hearts in tune\nEvery pattern finds its way\nThrough the hush before the day\n\n[Chorus]\nWe rise, we rise on silver strings\nCarry the pulse that morning brings\nWe rise, we rise, the sky our seam\nWoven together inside one dream\n\n[Bridge]\nClick of the shuttle, drum of the chest\nVoices answer east and west\nHold the line and let it gleam\nHold the line and keep the dream\n\n[Final Chorus]\nWe rise, we rise on silver strings\nCarry the pulse that morning brings\nWe rise, we rise, the sky our seam\nWoven together inside one dream'
  }
})
import asyncio
import os

from runware import Runware


async def main():
    async with Runware(api_key=os.environ["RUNWARE_API_KEY"]) as client:
        results = await client.run({
            "model": "runware:ace-step@v1.5-xl-base",
            "positivePrompt": "Three-minute cinematic electro-acoustic cantata inspired by a skyborne weaving guild gliding above green canyons. Start with close-mic textile loom clicks used as percussion, soft bowed vibraphone, and distant frame drums. Build into a 6/8 groove with kora arpeggios, warm analog bass, crisp handclaps, brushed cymbals, and airy modular synth pulses. Add a clear lead vocal with layered call-and-response choir, hopeful and adventurous mood, spacious production, wide stereo field, dynamic arrangement with verse, lift, chorus, instrumental bridge, final chorus, and gentle coda. Make the hook memorable, rhythmic, and singable, with natural human phrasing and no harsh distortion.",
            "negativePrompt": "muddy mix, clipped vocals, off-key singing, random noise, chaotic drums, weak bass, spoken narration, long silence, abrupt ending, distorted mastering, generic stock loop feel",
            "duration": 180,
            "seed": 64708,
            "steps": 50,
            "CFGScale": 9.5,
            "settings": {
                "bpm": 132,
                "keyScale": "D minor",
                "timeSignature": 6,
                "guidanceType": "apg",
                "cfgIntervalStart": 0,
                "cfgIntervalEnd": 0.85,
                "vocalLanguage": "en",
                "lyrics": "[Verse 1]\nThread by thread we cross the blue\nHands in rhythm, hearts in tune\nEvery pattern finds its way\nThrough the hush before the day\n\n[Chorus]\nWe rise, we rise on silver strings\nCarry the pulse that morning brings\nWe rise, we rise, the sky our seam\nWoven together inside one dream\n\n[Bridge]\nClick of the shuttle, drum of the chest\nVoices answer east and west\nHold the line and let it gleam\nHold the line and keep the dream\n\n[Final Chorus]\nWe rise, we rise on silver strings\nCarry the pulse that morning brings\nWe rise, we rise, the sky our seam\nWoven together inside one dream"
            }
        })


asyncio.run(main())
curl https://api.runware.ai/v1 \
  -H "Authorization: Bearer $RUNWARE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '[
    {
      "taskType": "audioInference",
      "taskUUID": "63675dd8-9c96-4d90-9eea-e239ce58d438",
      "model": "runware:ace-step@v1.5-xl-base",
      "positivePrompt": "Three-minute cinematic electro-acoustic cantata inspired by a skyborne weaving guild gliding above green canyons. Start with close-mic textile loom clicks used as percussion, soft bowed vibraphone, and distant frame drums. Build into a 6/8 groove with kora arpeggios, warm analog bass, crisp handclaps, brushed cymbals, and airy modular synth pulses. Add a clear lead vocal with layered call-and-response choir, hopeful and adventurous mood, spacious production, wide stereo field, dynamic arrangement with verse, lift, chorus, instrumental bridge, final chorus, and gentle coda. Make the hook memorable, rhythmic, and singable, with natural human phrasing and no harsh distortion.",
      "negativePrompt": "muddy mix, clipped vocals, off-key singing, random noise, chaotic drums, weak bass, spoken narration, long silence, abrupt ending, distorted mastering, generic stock loop feel",
      "duration": 180,
      "seed": 64708,
      "steps": 50,
      "CFGScale": 9.5,
      "settings": {
        "bpm": 132,
        "keyScale": "D minor",
        "timeSignature": 6,
        "guidanceType": "apg",
        "cfgIntervalStart": 0,
        "cfgIntervalEnd": 0.85,
        "vocalLanguage": "en",
        "lyrics": "[Verse 1]\nThread by thread we cross the blue\nHands in rhythm, hearts in tune\nEvery pattern finds its way\nThrough the hush before the day\n\n[Chorus]\nWe rise, we rise on silver strings\nCarry the pulse that morning brings\nWe rise, we rise, the sky our seam\nWoven together inside one dream\n\n[Bridge]\nClick of the shuttle, drum of the chest\nVoices answer east and west\nHold the line and let it gleam\nHold the line and keep the dream\n\n[Final Chorus]\nWe rise, we rise on silver strings\nCarry the pulse that morning brings\nWe rise, we rise, the sky our seam\nWoven together inside one dream"
      }
    }
  ]'
runware run runware:ace-step@v1.5-xl-base \
  positivePrompt="Three-minute cinematic electro-acoustic cantata inspired by a skyborne weaving guild gliding above green canyons. Start with close-mic textile loom clicks used as percussion, soft bowed vibraphone, and distant frame drums. Build into a 6/8 groove with kora arpeggios, warm analog bass, crisp handclaps, brushed cymbals, and airy modular synth pulses. Add a clear lead vocal with layered call-and-response choir, hopeful and adventurous mood, spacious production, wide stereo field, dynamic arrangement with verse, lift, chorus, instrumental bridge, final chorus, and gentle coda. Make the hook memorable, rhythmic, and singable, with natural human phrasing and no harsh distortion." \
  negativePrompt="muddy mix, clipped vocals, off-key singing, random noise, chaotic drums, weak bass, spoken narration, long silence, abrupt ending, distorted mastering, generic stock loop feel" \
  duration=180 \
  seed=64708 \
  steps=50 \
  CFGScale=9.5 \
  settings.bpm=132 \
  settings.keyScale="D minor" \
  settings.timeSignature=6 \
  settings.guidanceType=apg \
  settings.cfgIntervalStart=0 \
  settings.cfgIntervalEnd=0.85 \
  settings.vocalLanguage=en \
  settings.lyrics="[Verse 1]
Thread by thread we cross the blue
Hands in rhythm, hearts in tune
Every pattern finds its way
Through the hush before the day

[Chorus]
We rise, we rise on silver strings
Carry the pulse that morning brings
We rise, we rise, the sky our seam
Woven together inside one dream

[Bridge]
Click of the shuttle, drum of the chest
Voices answer east and west
Hold the line and let it gleam
Hold the line and keep the dream

[Final Chorus]
We rise, we rise on silver strings
Carry the pulse that morning brings
We rise, we rise, the sky our seam
Woven together inside one dream"
{
  "taskType": "audioInference",
  "taskUUID": "63675dd8-9c96-4d90-9eea-e239ce58d438",
  "model": "runware:ace-step@v1.5-xl-base",
  "positivePrompt": "Three-minute cinematic electro-acoustic cantata inspired by a skyborne weaving guild gliding above green canyons. Start with close-mic textile loom clicks used as percussion, soft bowed vibraphone, and distant frame drums. Build into a 6/8 groove with kora arpeggios, warm analog bass, crisp handclaps, brushed cymbals, and airy modular synth pulses. Add a clear lead vocal with layered call-and-response choir, hopeful and adventurous mood, spacious production, wide stereo field, dynamic arrangement with verse, lift, chorus, instrumental bridge, final chorus, and gentle coda. Make the hook memorable, rhythmic, and singable, with natural human phrasing and no harsh distortion.",
  "negativePrompt": "muddy mix, clipped vocals, off-key singing, random noise, chaotic drums, weak bass, spoken narration, long silence, abrupt ending, distorted mastering, generic stock loop feel",
  "duration": 180,
  "seed": 64708,
  "steps": 50,
  "CFGScale": 9.5,
  "settings": {
    "bpm": 132,
    "keyScale": "D minor",
    "timeSignature": 6,
    "guidanceType": "apg",
    "cfgIntervalStart": 0,
    "cfgIntervalEnd": 0.85,
    "vocalLanguage": "en",
    "lyrics": "[Verse 1]\nThread by thread we cross the blue\nHands in rhythm, hearts in tune\nEvery pattern finds its way\nThrough the hush before the day\n\n[Chorus]\nWe rise, we rise on silver strings\nCarry the pulse that morning brings\nWe rise, we rise, the sky our seam\nWoven together inside one dream\n\n[Bridge]\nClick of the shuttle, drum of the chest\nVoices answer east and west\nHold the line and let it gleam\nHold the line and keep the dream\n\n[Final Chorus]\nWe rise, we rise on silver strings\nCarry the pulse that morning brings\nWe rise, we rise, the sky our seam\nWoven together inside one dream"
  }
}
Response
{
  "taskType": "audioInference",
  "taskUUID": "63675dd8-9c96-4d90-9eea-e239ce58d438",
  "audioUUID": "27983512-0418-4067-b345-ee3b826a717a",
  "audioURL": "https://am.runware.ai/audio/os/a06dlim3/ws/5/ai/27983512-0418-4067-b345-ee3b826a717a.mp3",
  "seed": 64708,
  "cost": 0.0183
}
Text to Audio

Volcanic Trainyard Tango Suite

0:00
import { createClient } from '@runware/sdk'

const client = await createClient({ apiKey: process.env.RUNWARE_API_KEY })
await client.connect()

const [result] = await client.run({
  model: 'runware:ace-step@v1.5-xl-base',
  positivePrompt: 'A 150-second cinematic electro-tango set in a volcanic rail depot at midnight: bandoneon lead, muted trumpet replies, deep upright bass, brushed snare, crisp palmas, low analog synth pulse, occasional steel-wheel percussion, and a smoky solo female Spanish vocal. Structure: tense 8-bar intro with distant train brakes and sub-bass rumble, sensual verse, explosive pre-chorus with claps, hook with soaring bandoneon and trumpet, instrumental break using rail-metal hits as rhythm, final chorus with layered harmonies, clean resolved ending. High-fidelity studio mix, wide stereo field, punchy but warm low end, dramatic dynamic swells, modern soundtrack polish, danceable yet mysterious.',
  negativePrompt: 'muddy mix, clipping, harsh sibilance, weak bass, off-beat percussion, flat dynamics, noisy vocal take, distorted master, random genre shifts, spoken narration, crowd noise',
  duration: 150,
  seed: 86583,
  steps: 50,
  CFGScale: 12,
  settings: {
    bpm: 124,
    keyScale: 'D minor',
    timeSignature: 4,
    vocalLanguage: 'es',
    guidanceType: 'apg',
    cfgIntervalStart: 0,
    cfgIntervalEnd: 0.9,
    lyrics: '[Verso]\nBajo la vía late el fuego,\nmi sombra cruza sin mirar,\nun tren de hierro lleva el miedo,\ny tu nombre vuelve a sonar.\n\n[Pre-Coro]\nTacón, carbón, chispa en la piel,\nla noche gira contra el andén.\n\n[Coro]\nDame la señal, déjame caer,\nentre humo y compás voy a renacer.\nDame la verdad, no la de papel,\nsi el suelo va a temblar, baila conmigo otra vez.\n\n[Puente]\nRojo el pulso, negro el motor,\nrompe la cadena del viejo dolor.\n\n[Coro Final]\nDame la señal, déjame caer,\nentre humo y compás voy a renacer.\nDame la verdad, no la de papel,\nsi el suelo va a temblar, baila conmigo otra vez.'
  }
})
import asyncio
import os

from runware import Runware


async def main():
    async with Runware(api_key=os.environ["RUNWARE_API_KEY"]) as client:
        results = await client.run({
            "model": "runware:ace-step@v1.5-xl-base",
            "positivePrompt": "A 150-second cinematic electro-tango set in a volcanic rail depot at midnight: bandoneon lead, muted trumpet replies, deep upright bass, brushed snare, crisp palmas, low analog synth pulse, occasional steel-wheel percussion, and a smoky solo female Spanish vocal. Structure: tense 8-bar intro with distant train brakes and sub-bass rumble, sensual verse, explosive pre-chorus with claps, hook with soaring bandoneon and trumpet, instrumental break using rail-metal hits as rhythm, final chorus with layered harmonies, clean resolved ending. High-fidelity studio mix, wide stereo field, punchy but warm low end, dramatic dynamic swells, modern soundtrack polish, danceable yet mysterious.",
            "negativePrompt": "muddy mix, clipping, harsh sibilance, weak bass, off-beat percussion, flat dynamics, noisy vocal take, distorted master, random genre shifts, spoken narration, crowd noise",
            "duration": 150,
            "seed": 86583,
            "steps": 50,
            "CFGScale": 12,
            "settings": {
                "bpm": 124,
                "keyScale": "D minor",
                "timeSignature": 4,
                "vocalLanguage": "es",
                "guidanceType": "apg",
                "cfgIntervalStart": 0,
                "cfgIntervalEnd": 0.9,
                "lyrics": "[Verso]\nBajo la vía late el fuego,\nmi sombra cruza sin mirar,\nun tren de hierro lleva el miedo,\ny tu nombre vuelve a sonar.\n\n[Pre-Coro]\nTacón, carbón, chispa en la piel,\nla noche gira contra el andén.\n\n[Coro]\nDame la señal, déjame caer,\nentre humo y compás voy a renacer.\nDame la verdad, no la de papel,\nsi el suelo va a temblar, baila conmigo otra vez.\n\n[Puente]\nRojo el pulso, negro el motor,\nrompe la cadena del viejo dolor.\n\n[Coro Final]\nDame la señal, déjame caer,\nentre humo y compás voy a renacer.\nDame la verdad, no la de papel,\nsi el suelo va a temblar, baila conmigo otra vez."
            }
        })


asyncio.run(main())
curl https://api.runware.ai/v1 \
  -H "Authorization: Bearer $RUNWARE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '[
    {
      "taskType": "audioInference",
      "taskUUID": "0666a337-7803-47ed-a95e-3adfd2e996fb",
      "model": "runware:ace-step@v1.5-xl-base",
      "positivePrompt": "A 150-second cinematic electro-tango set in a volcanic rail depot at midnight: bandoneon lead, muted trumpet replies, deep upright bass, brushed snare, crisp palmas, low analog synth pulse, occasional steel-wheel percussion, and a smoky solo female Spanish vocal. Structure: tense 8-bar intro with distant train brakes and sub-bass rumble, sensual verse, explosive pre-chorus with claps, hook with soaring bandoneon and trumpet, instrumental break using rail-metal hits as rhythm, final chorus with layered harmonies, clean resolved ending. High-fidelity studio mix, wide stereo field, punchy but warm low end, dramatic dynamic swells, modern soundtrack polish, danceable yet mysterious.",
      "negativePrompt": "muddy mix, clipping, harsh sibilance, weak bass, off-beat percussion, flat dynamics, noisy vocal take, distorted master, random genre shifts, spoken narration, crowd noise",
      "duration": 150,
      "seed": 86583,
      "steps": 50,
      "CFGScale": 12,
      "settings": {
        "bpm": 124,
        "keyScale": "D minor",
        "timeSignature": 4,
        "vocalLanguage": "es",
        "guidanceType": "apg",
        "cfgIntervalStart": 0,
        "cfgIntervalEnd": 0.9,
        "lyrics": "[Verso]\nBajo la vía late el fuego,\nmi sombra cruza sin mirar,\nun tren de hierro lleva el miedo,\ny tu nombre vuelve a sonar.\n\n[Pre-Coro]\nTacón, carbón, chispa en la piel,\nla noche gira contra el andén.\n\n[Coro]\nDame la señal, déjame caer,\nentre humo y compás voy a renacer.\nDame la verdad, no la de papel,\nsi el suelo va a temblar, baila conmigo otra vez.\n\n[Puente]\nRojo el pulso, negro el motor,\nrompe la cadena del viejo dolor.\n\n[Coro Final]\nDame la señal, déjame caer,\nentre humo y compás voy a renacer.\nDame la verdad, no la de papel,\nsi el suelo va a temblar, baila conmigo otra vez."
      }
    }
  ]'
runware run runware:ace-step@v1.5-xl-base \
  positivePrompt="A 150-second cinematic electro-tango set in a volcanic rail depot at midnight: bandoneon lead, muted trumpet replies, deep upright bass, brushed snare, crisp palmas, low analog synth pulse, occasional steel-wheel percussion, and a smoky solo female Spanish vocal. Structure: tense 8-bar intro with distant train brakes and sub-bass rumble, sensual verse, explosive pre-chorus with claps, hook with soaring bandoneon and trumpet, instrumental break using rail-metal hits as rhythm, final chorus with layered harmonies, clean resolved ending. High-fidelity studio mix, wide stereo field, punchy but warm low end, dramatic dynamic swells, modern soundtrack polish, danceable yet mysterious." \
  negativePrompt="muddy mix, clipping, harsh sibilance, weak bass, off-beat percussion, flat dynamics, noisy vocal take, distorted master, random genre shifts, spoken narration, crowd noise" \
  duration=150 \
  seed=86583 \
  steps=50 \
  CFGScale=12 \
  settings.bpm=124 \
  settings.keyScale="D minor" \
  settings.timeSignature=4 \
  settings.vocalLanguage=es \
  settings.guidanceType=apg \
  settings.cfgIntervalStart=0 \
  settings.cfgIntervalEnd=0.9 \
  settings.lyrics="[Verso]
Bajo la vía late el fuego,
mi sombra cruza sin mirar,
un tren de hierro lleva el miedo,
y tu nombre vuelve a sonar.

[Pre-Coro]
Tacón, carbón, chispa en la piel,
la noche gira contra el andén.

[Coro]
Dame la señal, déjame caer,
entre humo y compás voy a renacer.
Dame la verdad, no la de papel,
si el suelo va a temblar, baila conmigo otra vez.

[Puente]
Rojo el pulso, negro el motor,
rompe la cadena del viejo dolor.

[Coro Final]
Dame la señal, déjame caer,
entre humo y compás voy a renacer.
Dame la verdad, no la de papel,
si el suelo va a temblar, baila conmigo otra vez."
{
  "taskType": "audioInference",
  "taskUUID": "0666a337-7803-47ed-a95e-3adfd2e996fb",
  "model": "runware:ace-step@v1.5-xl-base",
  "positivePrompt": "A 150-second cinematic electro-tango set in a volcanic rail depot at midnight: bandoneon lead, muted trumpet replies, deep upright bass, brushed snare, crisp palmas, low analog synth pulse, occasional steel-wheel percussion, and a smoky solo female Spanish vocal. Structure: tense 8-bar intro with distant train brakes and sub-bass rumble, sensual verse, explosive pre-chorus with claps, hook with soaring bandoneon and trumpet, instrumental break using rail-metal hits as rhythm, final chorus with layered harmonies, clean resolved ending. High-fidelity studio mix, wide stereo field, punchy but warm low end, dramatic dynamic swells, modern soundtrack polish, danceable yet mysterious.",
  "negativePrompt": "muddy mix, clipping, harsh sibilance, weak bass, off-beat percussion, flat dynamics, noisy vocal take, distorted master, random genre shifts, spoken narration, crowd noise",
  "duration": 150,
  "seed": 86583,
  "steps": 50,
  "CFGScale": 12,
  "settings": {
    "bpm": 124,
    "keyScale": "D minor",
    "timeSignature": 4,
    "vocalLanguage": "es",
    "guidanceType": "apg",
    "cfgIntervalStart": 0,
    "cfgIntervalEnd": 0.9,
    "lyrics": "[Verso]\nBajo la vía late el fuego,\nmi sombra cruza sin mirar,\nun tren de hierro lleva el miedo,\ny tu nombre vuelve a sonar.\n\n[Pre-Coro]\nTacón, carbón, chispa en la piel,\nla noche gira contra el andén.\n\n[Coro]\nDame la señal, déjame caer,\nentre humo y compás voy a renacer.\nDame la verdad, no la de papel,\nsi el suelo va a temblar, baila conmigo otra vez.\n\n[Puente]\nRojo el pulso, negro el motor,\nrompe la cadena del viejo dolor.\n\n[Coro Final]\nDame la señal, déjame caer,\nentre humo y compás voy a renacer.\nDame la verdad, no la de papel,\nsi el suelo va a temblar, baila conmigo otra vez."
  }
}
Response
{
  "taskType": "audioInference",
  "taskUUID": "0666a337-7803-47ed-a95e-3adfd2e996fb",
  "audioUUID": "7652545e-1d19-481d-8423-67b64d92a708",
  "audioURL": "https://am.runware.ai/audio/os/a04d20/ws/5/ai/7652545e-1d19-481d-8423-67b64d92a708.mp3",
  "seed": 86583,
  "cost": 0.0183
}