# Text to Audio You can start an audio workflow from a text node. Write a script, describe a sound, or sketch the music you want, then connect it to an audio node and pick a model. What comes back is playable audio on the canvas. Ready to export, layer under a video, or drive a lipsynced performance in a video node. Speech Text-to-speech turns a written script into a voiceover. Type what you want said, pick a voice from the preset library, and generate. Good for narration on explainer videos, scratch VO for concept pitches, or any moment where a voice needs to land on the canvas without a recording session. You can interact with it in a number of ways: * Type or paste your script into the prompt field * Select a voice from the dropdown, or use a recorded custom voice asset * Use expressive prompt tags (e.g. whisper, shout, happy) on models that support them to shape delivery * Connect the output to a video node with a lipsync model to drive a lipsynced performance SFX Text-to-SFX generates individual sound effects from a plain-language description. Describe the sound — "heavy wooden door slamming shut in a stone hallway" — and generate. Useful for ambient texture, hit sounds, foley, and anywhere you'd otherwise be digging through a stock library. Best practices: * Be specific about the source and environment (material, space, distance) * Generate a few variations with different seeds and pick the strongest * Layer multiple SFX audio nodes for richer atmospheres Music Text-to-music generates short music tracks from a text prompt describing genre, mood, instrumentation, or reference. Useful for background beds on social clips, pitch videos, and mood-setting on ad concepts where you don't want to license a track. You can interact with it in a number of ways: * Describe the track — genre, tempo, instruments, mood, references * Generate variations across seeds to find the right feel * Pair music output with a video node for a scored clip How to use: Here are some example workflows using Text to Audio: * Write a script → generate speech → connect to a video node with a lipsync model → export a lipsynced clip * Generate a product-demo voiceover → layer a music bed from text-to-music → export a finished promo * Generate three SFX variations for a scene → pick the strongest → layer onto a generated video Models Visit our [Audio Models](/models/audio-models.md) section to learn about the speech, SFX, and music models available in the Audio Node. ## Limits & formats **Character limit (ElevenLabs Multilingual v2)** ElevenLabs Multilingual v2 has a 450-character limit per generation. Text longer than 450 characters is truncated — this is why a script with 800+ characters may only produce audio for the first \~450. To narrate longer scripts, split the text across multiple audio nodes and chain them. **Output format** Audio is delivered as MP3. WAV output is not currently available for ElevenLabs voice models in FLORA. **Bitrate** The MP3 bitrate is fixed at 128 kbps and cannot be changed. --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.flora.ai/nodes/audio-node/text-to-video.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.