genblaze-elevenlabs

ElevenLabs text-to-speech + sound-effects provider adapter for genblaze — AI voice and SFX pipelines with SHA-256 provenance manifests on every clip.

genblaze-elevenlabs wraps ElevenLabs' TTS models (Eleven v3, Multilingual v2, Flash v2.5) and text-to-sound-effects API as genblaze providers. Generate narration, dialogue, and sound design with tracked provenance; compose into multi-step AI pipelines; persist outputs to Backblaze B2 or any S3-compatible store.

Why genblaze-elevenlabs

• Eleven v3 + Multilingual v2 + Flash v2.5 — Pick quality (v3), breadth (multilingual, 29 languages), or latency (~75 ms for Flash).
• Text-to-sound-effects — generate Foley / ambient sound with the ElevenLabsSFXProvider.
• Provenance by default — SHA-256-verified manifest on every clip; embed manifest directly into MP3 files.
• Word-level timings — ElevenLabs alignment flows into genblaze's WordTiming / Track models for downstream subtitling + AV compositing.
• Production-ready — retries, timeouts, moderation hooks, step caching.
• Durable storage — plug genblaze-s3 in for Backblaze B2 / AWS S3 / R2 / MinIO persistence.

Providers + models

Registered via entry points as elevenlabs-tts and elevenlabs-sfx.

Install

pip install genblaze-elevenlabs

Quickstart — TTS narration

pip install genblaze-core genblaze-elevenlabs
export ELEVENLABS_API_KEY="..."

from genblaze_core import Modality, Pipeline
from genblaze_elevenlabs import ElevenLabsTTSProvider

run, manifest = (
    Pipeline("narration")
    .step(ElevenLabsTTSProvider(output_dir="output/audio"),
          model="eleven_v3",
          prompt="Welcome to Genblaze — AI media pipelines with full provenance.",
          modality=Modality.AUDIO,
          voice_id="JBFqnCBsd6RMkjVDRZzb",
          output_format="mp3_44100_128")
    .run(timeout=60)
)
print(run.steps[0].assets[0].url, manifest.canonical_hash)

Quickstart — text-to-sound-effects

from genblaze_elevenlabs import ElevenLabsSFXProvider

run, manifest = (
    Pipeline("sfx")
    .step(ElevenLabsSFXProvider(output_dir="output/sfx"),
          model="eleven_text_to_sound_v2",
          prompt="Thunder crashing during a heavy rainstorm with distant rumbling",
          modality=Modality.AUDIO, duration_seconds=10)
    .run(timeout=60)
)

Persist to Backblaze B2

from genblaze_core import KeyStrategy, ObjectStorageSink
from genblaze_s3 import S3StorageBackend

storage = ObjectStorageSink(
    S3StorageBackend.for_backblaze("my-bucket"),
    key_strategy=KeyStrategy.HIERARCHICAL,
)
# pass sink=storage to .run(…)

Backblaze B2 is the recommended default sink for AI-generated audio — cost-efficient, S3-compatible, Object Lock for immutable manifests.

Credentials

Documentation

• Main repo: https://github.com/backblaze-labs/genblaze
• Examples: elevenlabs_tts_pipeline.py · elevenlabs_sfx_pipeline.py

Related packages

• genblaze-core — the pipeline SDK
• genblaze-s3 — durable storage on Backblaze B2 and other S3-compatible backends
• Other audio providers: genblaze-openai (TTS) · genblaze-lmnt (fast TTS) · genblaze-stability-audio (music) · genblaze-gmicloud

License

MIT