--- license: apache-2.0 library_name: inworld-sava pipeline_tag: voice-activity-detection tags: [audio, speech, voice-activity-detection, vad, semantic-vad, turn-detection, end-of-turn, streaming, realtime] ---

Inworld

SAVA

Semantic acoustic voice activity: streaming speech and end-of-turn detection in one model with an open license.

GitHub: inworld-ai/sava

--- ## TL;DR SAVA returns two probabilities for a live audio stream: * `speech` → whether a frame holds speech * `turn` → whether the speaker's turn has ended The model's shared network uses acoustic and sentence context to distinguish speech from coughs, laughs, music and keyboard clatter. | | SAVA | |---|---| | made for | voice activity and conversational turn taking | | input | 16 kHz mono audio | | output | two probabilities every 40 ms | | timing | speech probability for a frame is emitted 120 ms later (lookahead); end-of-turn probability has no lookahead | | parameters | 25.5M | | execution | plain PyTorch, CPU or an explicitly selected device | Use it to open a turn when speech starts and close it when the sentence is complete. The [reference turn policy](https://github.com/inworld-ai/sava/blob/main/docs/USAGE.md) turns the probabilities into segments; applications can supply their own policy. SAVA responds to any voice. It does not pick out a target speaker, and it leaves interruptions and backchannels to the application's dialogue policy. ## Quick start Using a coding agent? Point it at the [setup section of AGENTS.md](https://github.com/inworld-ai/sava/blob/main/AGENTS.md#setting-up-sava-for-a-user): it installs the package, checks the machine with `sava-doctor` and runs SAVA on a first recording. With Python 3.10 or later: ```bash python -m pip install inworld-sava # Or from a clone: git clone https://github.com/inworld-ai/sava && cd sava && python -m pip install -e . sava run recording.wav sava run recording.wav --turns_dir turns/ --curves curves.json ``` ```python from inworld import sava model = sava.load("inworld/sava") curves = sava.run(model, sava.load_audio("recording.wav")) turns = sava.segment(curves, sava.Policy(), model.lookahead_frames) session = sava.Session(model) for piece in pieces_of_16k_audio: # One-dimensional floating-point mono samples. frames = session.push(piece) frames = session.flush() ``` `sava.load("inworld/sava")` and the CLI's default model download the weights on first use and verify their checksum. The CLI defaults to CPU; `--device` selects another device. For microphone input, install `python -m pip install "inworld-sava[mic]"` and run `sava mic`. The code, [the integration guide](https://github.com/inworld-ai/sava/blob/main/docs/USAGE.md) and the [benchmarks](https://github.com/inworld-ai/sava/blob/main/bench/README.md) are on GitHub: [inworld-ai/sava](https://github.com/inworld-ai/sava). ## Hardware Tested on NVIDIA Ampere, Hopper and Blackwell GPUs and on CPU. On an Apple M3 Max CPU one stream takes 3.8 ms per 40 ms frame; [the integration guide](https://github.com/inworld-ai/sava/blob/main/docs/USAGE.md) lists the measured configurations. ## The benchmarks

Voice-activity frame AUC for SAVA and Silero VAD on nine public dataset readings

End-of-turn clip AUC for SAVA and smart-turn on public datasets

Voice activity is compared with Silero VAD; end of turn is compared with pipecat smart-turn v3.2. These are per-dataset measurements, with no combined ranking. Speech scores use each model's own output timing, so SAVA's 120 ms lookahead makes the comparison unequal in latency. Quiet speech-like noise can trigger SAVA; the ESC-50 results measure that failure mode. [The complete tables, confidence intervals, noise tests, turn latency and reproduction steps](https://github.com/inworld-ai/sava/blob/main/bench/README.md) are on GitHub. ## Architecture

SAVA: causal log-mel features, frame mixer, Mamba and attention trunk, speech and turn heads

| part | detail | |---|---| | front end | causal log-mel, 128 Slaney mel bins of a 25 ms Hann DFT every 10 ms, log10 with a 1e-10 floor, `(x + 4) / 4`; four hops per frame plus the frame's log RMS | | mixer | the current and previous frame's features, one linear projection to 512, RMSNorm | | trunk | three Mamba-2 blocks (expand 2, head dim 64, state 128, conv kernel 4) and one attention block (8 query heads, 2 key/value heads, head dim 64, 512-frame causal window, rotary positions, RMS-normed q and k, exclusive self attention), each with a SwiGLU FFN of 2048 | | voice-activity head | RMSNorm, down-projection to 512, one attention block (window 512, FFN 1536), RMSNorm, two linear readouts: on time, and the released three-frame look-ahead | | end-of-turn head | the trunk output with a 64-wide embedding of the previous frame's on-time speech probability, the same block shape, a linear readout with four horizon columns (0, 200, 400, 800 ms); column 0 is the released output | ## Training data SAVA was trained on approximately 5,000 hours of unique audio. Inworld built training examples from real and synthetic conversational speech mixed with non-verbal vocal sounds, noise, music and room acoustics, with a speech label and an end-of-turn label for every frame. The training recipe, code and data are not part of this release. Third-party datasets used in training are credited in [NOTICE](NOTICE). ## License Code and weights are Apache-2.0 ([LICENSE](LICENSE)). Third-party notices are in [NOTICE](NOTICE). ## How to contribute Everyone is welcome to use the model however they like and to make it better in any way they see fit! Development happens on GitHub: see [CONTRIBUTING.md](https://github.com/inworld-ai/sava/blob/main/CONTRIBUTING.md) for setup, tests and bug reports. ---

Built for people with ❤️ by Inworld.