Text-to-speech

The MediaPipe Text-to-Speech task lets you synthesize natural-sounding speech audio from text directly on-device using the open-weight Kokoro-82M neural text-to-speech model packaged in a single .litertlm container. The task operates as a headless PCM audio generator, returning normalized 24 kHz mono 32-bit floating-point (float32) audio samples that your application can play through platform audio APIs, save as WAV files, or pass to downstream audio pipelines.

• Input: UTF-8 text string (single sentence, multi-sentence paragraph, or incremental text deltas pushed using TtsStream) and optional per-utterance SynthesisOptions.
• Output: AudioResult (unary) or a stream of AudioChunk buffers (streaming) containing normalized 32-bit float mono PCM samples in [-1.0, 1.0] at 24,000 Hz.

Key capabilities of the MediaPipe Text-to-Speech task with the Kokoro-82M backend include:

• Single-file .litertlm deployment: Packages the acoustic generator, iSTFT neural vocoder, phonetic dictionaries for all supported languages, and dozens of voice style embeddings into a single memory-mapped .litertlm bundle.
• Low-latency streaming speech synthesis: Streams AudioChunk buffers clause-by-clause (synthesizeStream / synthesize_stream) as each clause finishes vocoding so audio playback starts after the first clause, and supports incremental token-by-token input (openStream / open_stream with pushText and finishInput) for real-time LLM-to-TTS pipelines.
• Sub-millisecond multilingual voice switching: Switch between any of the bundled voices on a per-utterance basis using SynthesisOptions in < 1 ms without reloading model weights.
• Automatic sentence chunking: Splits multi-sentence paragraphs at clause and sentence boundaries (., !, ?, ;, :, \n) so long inputs stay within the model's context window.
• Immediate barge-in cancellation: Thread-safe, lock-free cancel() API aborts in-flight unary or streaming synthesis within milliseconds when a user interrupts or a coroutine, async sequence, or task is canceled.

Note: Please ensure you process voice data responsibly and in accordance with applicable law.

Get Started

Start using this task by following one of these implementation guides for your target platform. These platform-specific guides walk you through a basic implementation of this task, including a recommended model, and code example with recommended configuration options:

Platform	Code Example	Guide
Android	Code example	Guide
iOS	Code example	Guide

How it works

The MediaPipe Text-to-Speech task executes Kokoro-82M through a multi-stage runtime pipeline:

1. Sentence segmentation (ExtractNextChunk): Multi-sentence input text (or incremental text pushed using TtsStream) is segmented at punctuation boundaries (max_buffer_size = 200 characters) to fit the acoustic model's token capacity.
2. Grapheme-to-Phoneme (G2P) frontend: Converts normalized text into phoneme token IDs using the phonetic dictionaries bundled in the model for each supported language.
3. Acoustic stage (KokoroAcousticStage): Runs the LiteRT acoustic model conditioned on the selected voice's 510 x 256 style embedding vector to predict duration tokens, pitch (F0) and energy contours, and 512-channel acoustic features.
4. Neural vocoder stage (KokoroVocoderStage): Runs the ISTFTNet-based LiteRT vocoder model (accelerated on CPU using XNNPACK or hybrid CPU+GPU) to convert the acoustic features and contours into a 24 kHz mono float32 PCM waveform. During streaming synthesis, each clause's vocoded AudioChunk is emitted immediately with MultiStagedSession::ProcessNext(), while unary synthesis drains the stream into a single concatenated AudioResult.

Models

The MediaPipe Text-to-Speech task requires a speech synthesis model in the LiteRT-LM (.litertlm) format. Models aren't included in the MediaPipe Tasks libraries. Download the pre-packaged kokoro_int8_gpu.litertlm model bundle from Hugging Face, store it on the device, and pass the local path to BaseOptions.model_asset_path.

Model Bundle	Size	Quantization	License	Supported Locales	Recommended Use Case
kokoro_int8_gpu.litertlm	~351 MB	INT8 acoustic + INT8 GPU vocoder	Apache 2.0	All nine locales (en-US, en-GB, es, fr-FR, it, pt-BR, hi, cmn, ja)	Multilingual on-device speech synthesis across mobile, desktop, and web.

Supported languages and voices

Kokoro-82M includes 54 pre-packaged voice embeddings across nine locales (eight languages). Each voice ID uses a two-letter prefix <lang_char><gender_char>_<name> (for example, af_heart = American English Female "Heart", ef_dora = Spanish Female "Dora").

Note: In Kokoro-82M, the voice prefix determines the phonemizer language. When switching voices using SynthesisOptions, you only need to set voice_id—leave locale unset (null / None) so the task automatically infers the matching BCP 47 language tag from the voice ID.

Language	BCP 47 Language Tag	Prefix	Bundled Voice IDs (54 Total)
English (American)	en-US	af_, am_	Female (11): af_alloy, af_aoede, af_bella, af_heart (default), af_jessica, af_kore, af_nicole, af_nova, af_river, af_sarah, af_sky Male (9): am_adam, am_echo, am_eric, am_fenrir, am_liam, am_michael, am_onyx, am_puck, am_santa
English (British)	en-GB	bf_, bm_	Female (4): bf_alice, bf_emma, bf_isabella, bf_lily Male (4): bm_daniel, bm_fable, bm_george, bm_lewis
Spanish	es	ef_, em_	Female (1): ef_dora Male (2): em_alex, em_santa
French	fr-FR	ff_	Female (1): ff_siwis
Hindi	hi	hf_, hm_	Female (2): hf_alpha, hf_beta Male (2): hm_omega, hm_psi
Italian	it	if_, im_	Female (1): if_sara Male (1): im_nicola
Portuguese (Brazilian)	pt-BR	pf_, pm_	Female (1): pf_dora Male (2): pm_alex, pm_santa
Mandarin Chinese	zh-CN / cmn	zf_, zm_	Female (4): zf_xiaobei, zf_xiaoni, zf_xiaoxiao, zf_xiaoyi Male (4): zm_yunjian, zm_yunxi, zm_yunxia, zm_yunyang
Japanese	ja	jf_, jm_	Female (4): jf_alpha, jf_gongitsune, jf_nezumi, jf_tebukuro Male (1): jm_kumo

You can enumerate all voices present in the loaded .litertlm bundle at runtime using listVoices() / list_voices().

Task configuration

Task options (TextToSpeechOptions)

Option Name	Description	Value Range	Default Value
base_options.model_asset_path	Path (or URL on Web) to the single .litertlm model bundle containing the Kokoro-82M acoustic and vocoder graphs, phonetic lexicons, and voice embeddings.	String / Path	Required
base_options.delegate	Hardware acceleration delegate used by the LiteRT runtime. When set to GPU, the Kokoro acoustic stage executes on CPU (XNNPACK) and the neural vocoder executes on GPU (OpenCL/Metal/WebGPU) with CPU fallback.	BaseOptions.Delegate.CPU, BaseOptions.Delegate.GPU	BaseOptions.Delegate.CPU
speaking_rate	Speaking rate multiplier applied to synthesized Kokoro audio (1.0 represents normal speaking speed).	Float (0.1, 4.0]	1.0
target_bucket	Target sequence-length bucket for the Kokoro acoustic/vocoder graph (for example, 64, 128, or 256 tokens). Smaller buckets reduce synthesis latency for short sentences; 0 uses the model's default capacity.	0, 64, 128, 256	0
default_voice_id	Default Kokoro voice profile identifier (for example, "af_heart", "ef_dora", or "zf_xiaobei"). Can be overridden per utterance using SynthesisOptions.	String (Optional)	"af_heart"
default_locale	Optional default BCP 47 language tag (for example, "en-US", "es", "zh-CN"). When omitted, the task automatically infers the language from the selected Kokoro voice prefix.	BCP 47 String (Optional)	Inferred from voice
cache_directory	Writable directory path used to persist unpacked phonetic dictionary assets and LiteRT XNNPACK weight caches across runs for faster cold starts.	Directory Path (Optional)	""

Per-utterance options (SynthesisOptions)

Option Name	Description	Default Value
voice_id	Overrides the voice profile for a single synthesize call (for example, switching from "af_heart" to "ef_dora" in under 1 ms without reloading model weights).	null / None (uses task default)
locale	Optional BCP 47 locale override for a single utterance. Leave unset (null / None) when switching voices so the engine automatically resolves the voice's target language from its prefix.	null / None (inferred from voiceId)