Higgsfield AI said it has made ByteDance's all-in-one audio generation model, Seed Audio 1.0, available on its own platform and inside Claude via the Higgsfield MCP. The model supports text-to-narration, voice changing, and dubbing videos into 18 languages.
June 30, 2026 · Higgsfield AI
One Model for Voice, Music, FX & Ambience
Higgsfield folds ByteDance's Seed Audio 1.0 into its creative suite — a multimodal audio model that generates speech, sound effects and music in a single pass, with dubbing across 18 languages and direct access from Claude via MCP.
$0.1875
per generated minute (fal.ai API)
2 min
audio per request · 2,048-char script
18
languages for dubbing in Higgsfield
50+
languages cited via the Claude MCP link
Single-pass generation
Four audio layers, one request
vs
Conventional TTS
Voice — speech & narrationSound effects & foleyMusic bedsAmbient sound
Script to voiced video, inside one agent
Claudevia MCP
→
HiggsfieldVoiceover · Change Voice · Translate
→
Seed Audio 1.0generated audio
What testers praise
"Human-like" emotional inflection & voice consistency
Voice acting & foley said to top Seedance 2's native audio
Affordable enough for free experimentation
Strong for vlogs, narration & promo video
Where it falls short
Second voice in dialogue can sound mechanical
Multi-speaker generation can be inconsistent
Longer clips & intonation need more testing
Quality depends heavily on the prompt
Also supports zero-shot voice cloning from a short clip · multi-speaker generation · cross-lingual synthesis · text, reference-audio & image inputs. Available via Higgsfield, fal.ai API, Volcano Engine / BytePlus, and Claude + MCP.
Continue reading The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.
Already purchased? Sign in ✓ Signed in — this article isn’t included in your current plan.Unlocking the full article…