Sonilo¶
Development line: project:sonilo · thread sonilo-v1-1
Last event: 2026-06-29 · 1 dated since 2026-06-29 · Researched: 2026-09-05 · confidence: medium
What it is¶
Sonilo is an API-first audio tool for editors, creative-product teams, and developers who need music or sound tied to a cut.
- generates music from text or video;
- generates text- and video-led sound effects;
- mixes supplied voice and music with automatic ducking.
Development line¶
- 2026-06-29 — Sonilo v1.1 linked to a text-to-music endpoint. It delivers tighter video-to-music alignment, preserves original speech, and adds per-segment prompts. The first-party release itself is dated 2026-06-18.
What changed¶
2026-06-29 — Sonilo v1.1 delivers tighter video-to-music alignment, preserves original speech, and adds per-segment prompts. The first-party release itself is dated 2026-06-18.
How to use this¶
From 2026-06-29, we treat Sonilo v1.1 as a distinct candidate endpoint to evaluate through the linked text-to-music route and Sonilo platform, without assuming capabilities or migration requirements until the source pages are researched.
- Create an API key, keep it server-side, and add the required API balance before attempting generation. — https://platform.sonilo.com/docs/quickstart
- Preflight the account’s enabled services and usage, then select text-to-music for a musical brief or video-to-music for a locked visual timeline. — https://platform.sonilo.com/docs/api
- For direct text generation, POST a prompt and a 5–360 second duration to /v1/text-to-music. — https://platform.sonilo.com/docs/text-to-music
- Consume streamed audio chunks only through the complete event, or use async mode and poll the returned task ID to a terminal state. — https://platform.sonilo.com/docs/text-to-music
- If the product already uses fal, call the v1.1 text model at sonilo/v1.1/text-to-music rather than building a separate provider integration. — https://fal.ai/models/sonilo/v1.1/text-to-music
Best practices¶
- Keep API keys in a server-side environment variable. Revoke and replace an exposed key rather than embedding it in a client app. — https://platform.sonilo.com/docs
- Treat 401, 402, 403, and 429 as distinct setup, balance, service-access, and backoff conditions. Do not create placeholder audio after a failed request. — https://platform.sonilo.com/docs/text-to-music
- Use video input when timing, cuts, or voice space must guide the soundtrack. Use text input when the task is a standalone cue. — https://platform.sonilo.com/docs/api
- For paid or client work, review the generated track against voice-over and scene changes, retain the output and account record, and confirm that the current plan covers the intended release. — https://sonilo.com/blog/ai-music-video-creators-2026
Superseded by this¶
- 2026-06-18 — For video scoring that requires preserved dialogue or scene-specific direction, v1.1 supersedes v1.0-style generation without those controls. The release does not state that v1.0 endpoints were retired.
- 2026-08-03 — Where a project already has separate voice and music assets, the documented audio-ducking endpoint supersedes manual volume riding when its two-input workflow fits.
Still unknown¶
- The supplied event is dated 2026-06-29, while Sonilo’s release page dates v1.1 to 2026-06-18; no primary source found explains the eleven-day difference.
- The v1.1 comparison figures are vendor-reported and no public evaluation protocol or independent replication was found.
- The direct Sonilo text endpoint documents a 360-second ceiling, while the fal v1.1 text endpoint documents 600 seconds; treat limits as route-specific until confirmed for the intended account.
- No public primary source found in this pass declares a successor model after v1.1.
Sources¶
| source | title | read |
|---|---|---|
| https://sonilo.com/news/sonilo-v1-1 | Introducing Sonilo v1.1 | Sonilo |
| https://fal.ai/models/sonilo/v1.1/text-to-music | Sonilo V1.1 Text to Music (Text to Audio) API on fal | 2026-09-05 |
| https://platform.sonilo.com/ | Sonilo API — Enterprise AI Music API | 2026-09-05 |
| https://platform.sonilo.com/docs | Introduction · Sonilo Docs | 2026-09-05 |
| https://platform.sonilo.com/docs/quickstart | Quickstart · Sonilo Docs | 2026-09-05 |
| https://platform.sonilo.com/docs/text-to-music | Text to Music · Sonilo Docs | 2026-09-05 |
| https://platform.sonilo.com/docs/api | API reference | 2026-09-05 |
| https://sonilo.com/zh-CN/news/sonilo-v1 | 介绍 Sonilo v1.0 | Sonilo |
| https://sonilo.com/blog/sonilo-comfyui | Exciting News: Sonilo is Now Live on ComfyUI! | Sonilo |
| https://sonilo.com/blog/product/audio-ducking-api-guide | Audio Ducking API: Automatically Mix Voice and Music | Sonilo |
| https://sonilo.com/blog/ai-music-video-creators-2026 | Licensed AI Music for YouTube, Ads, and Client Videos (2026 Guide) | Sonilo |
Agent brief¶
- Subject:
project:sonilo, threadsonilo-v1-1, 1 dated events 2026-06-29 → 2026-06-29. - Practical note: From 2026-06-29, we treat Sonilo v1.1 as a distinct candidate endpoint to evaluate through the linked text-to-music route and Sonilo platform, without assuming capabilities or migration requirements until the source pages are researched.
- Confidence: medium. Dated supersedes above are the authority for what is obsolete.