ACE-Step 1.5XL SFT： Text to Music

# Python 3.10+: pip install comfy-sdk==0.4.0
import os

from comfy_sdk import Comfy

api_key = os.environ.get("COMFY_API_KEY", "").strip()
if not api_key: raise RuntimeError("Set COMFY_API_KEY")
os.environ.setdefault("COMFY_BASE_URL", "https://cloud.comfy.org")
client = Comfy(api_key=api_key)
workflow = client.workflows.from_json({
    "3": {
        "class_type": "KSampler",
        "inputs": {
            "cfg": 7,
            "denoise": 1,
            "latent_image": [
                "98",
                0
            ],
            "model": [
                "78",
                0
            ],
            "negative": [
                "47",
                0
            ],
            "positive": [
                "94",
                0
            ],
            "sampler_name": "euler",
            "scheduler": "simple",
            "seed": [
                "109",
                0
            ],
            "steps": 50
        },
        "_meta": {
            "title": "KSampler"
        }
    },
    "18": {
        "class_type": "VAEDecodeAudio",
        "inputs": {
            "samples": [
                "3",
                0
            ],
            "vae": [
                "106",
                0
            ]
        },
        "_meta": {
            "title": "VAE Decode Audio"
        }
    },
    "47": {
        "class_type": "ConditioningZeroOut",
        "inputs": {
            "conditioning": [
                "94",
                0
            ]
        },
        "_meta": {
            "title": "Conditioning Zero Out"
        }
    },
    "78": {
        "class_type": "ModelSamplingAuraFlow",
        "inputs": {
            "model": [
                "104",
                0
            ],
            "sampling": "flow",
            "shift": 3
        },
        "_meta": {
            "title": "ModelSamplingAuraFlow"
        }
    },
    "94": {
        "class_type": "TextEncodeAceStepAudio1.5",
        "inputs": {
            "bpm": 85,
            "cfg_scale": 2,
            "clip": [
                "105",
                0
            ],
            "duration": 120,
            "generate_audio_codes": True,
            "keyscale": "E minor",
            "language": "en",
            "lyrics": "[Intro]\n\n[Verse 1]\nDust on the dashboard, sun in the glass\nMile after mile of the same yellow grass\nYou left a cassette with a song you never sang\nI play it at dawn when the long shadows hang\n\n[Chorus]\nOh, the road keeps turning, turning gold\nEvery motel sign is a story told\nI'll keep driving till the evening's old\nFollowing the light you left me\n\n[Verse 2]\nCoffee gone cold at a two-pump town\nRadio static when the hills come down\nI wrote your name on a map in the rain\nFaded by morning, I wrote it again\n\n[Chorus]\nOh, the road keeps turning, turning gold\nEvery motel sign is a story told\nI'll keep driving till the evening's old\nFollowing the light you left me\n\n[Bridge]\n\n[Outro]\nFollowing the light you left me",
            "min_p": 0,
            "seed": [
                "109",
                0
            ],
            "tags": "A sun-drenched, mid-tempo desert dream-pop ballad set at a relaxed eighty-five beats per minute, carried by a steady, dry drum groove with soft brush snares and an earthy upright bassline. Warm, chorus-drenched electric guitars interlock with dusty electric piano chords and shimmering Mellotron strings across an expansive stereo spread. The lead vocal is an intimate, breathy male falsetto, captured close-mic with gentle plate reverb and vintage tape flutter, delivering bittersweet nostalgia and open-road yearning. In the bridge, a mournful muted trumpet solo rises through lush analog synth swells, while the mix prioritizes analog saturation, crisp transient percussion, and an airy, golden-hour atmosphere with deep low-end warmth and wide spatial separation.",
            "temperature": 0.85,
            "timesignature": "4",
            "top_k": 0,
            "top_p": 1
        },
        "_meta": {
            "title": "TextEncodeAceStepAudio1.5"
        }
    },
    "98": {
        "class_type": "EmptyAceStep1.5LatentAudio",
        "inputs": {
            "batch_size": 1,
            "seconds": 120
        },
        "_meta": {
            "title": "Empty Ace Step 1.5 Latent Audio"
        }
    },
    "104": {
        "class_type": "UNETLoader",
        "inputs": {
            "unet_name": "acestep_v1.5_xl_sft_bf16.safetensors",
            "weight_dtype": "default"
        },
        "_meta": {
            "title": "Load Diffusion Model"
        }
    },
    "105": {
        "class_type": "DualCLIPLoader",
        "inputs": {
            "clip_name1": "qwen_0.6b_ace15.safetensors",
            "clip_name2": "qwen_4b_ace15.safetensors",
            "device": "default",
            "type": "ace"
        },
        "_meta": {
            "title": "Load CLIP (Dual)"
        }
    },
    "106": {
        "class_type": "VAELoader",
        "inputs": {
            "vae_name": "ace_1.5_vae.safetensors"
        },
        "_meta": {
            "title": "Load VAE"
        }
    },
    "107": {
        "class_type": "SaveAudioMP3",
        "inputs": {
            "audio": [
                "18",
                0
            ],
            "filename_prefix": "audio/shift6",
            "quality": "V0"
        },
        "_meta": {
            "title": "Save Audio (MP3) (DEPRECATED)"
        }
    },
    "109": {
        "class_type": "PrimitiveInt",
        "inputs": {
            "value": 0
        },
        "_meta": {
            "title": "Int"
        }
    }
})

# run() also passes api_key to the partner nodes in this workflow.
job = client.run(workflow, api_key=api_key)
for output in job.get_outputs("107"):
    output.to_file(output.name)