SoTA open-source TTS
This report summarizes scanner evidence for resemble-ai/chatterbox, a Python project with 26,834 GitHub stars. SynthScan examined 9,827 lines of code across 62 source files, recording 121 pattern matches distributed across 14 syntactic categories. The legacy adjusted pattern signal is 20.4 (Elevated pattern signal); it is not an authorship estimate.
Use the findings to ask whether maintainers can explain intent, verify dependencies, simplify unnecessary scaffolding, and focus review effort. Risk Payload v1 adds ownership, supply-chain, over-engineering, and review-effort vectors when available. These are prompts for human verification, not probabilities, security findings, or measured maintenance costs.
Use each vector to guide manual verification of the underlying findings.
Can maintainers explain intent, assumptions, and trade-offs? 62 finding(s).
Are imports declared and package/API names verified? 28 finding(s).
Are wrappers, guards, and repeated structures necessary? 53 finding(s).
Which changes need focused human review? 121 finding(s).
This chart tracks the legacy adjusted pattern score across scan runs. Changes can reflect source changes, scan coverage, or detector/scoring versions; they do not measure AI adoption or prove that a team refactored code. Compare runs only when their scoring versions are compatible. The dashed line tracks the raw pattern-hit count.
Groups findings by configured severity weight. Severity is not a confidence estimate, security rating, or proof that a package is missing. Verify imports against the project's manifests and inspect each finding in its local design context.
This horizontal bar chart shows where configured pattern scores accumulate across directories. Treat the bars as locations to inspect, not as evidence of authorship or code quality. Use the file-level findings to decide whether a directory needs a focused ownership, dependency, complexity, or review check.
The scanner identified 121 distinct pattern matches across 14 syntactic categories. Each entry below represents a discrete location in the source code where the engine recorded a detector match to inspect in context. Expand any category row to inspect the individual file paths, line numbers, code snippets, and the lexical context (CODE, COMMENT, or STRING) in which each match was detected.
Reading the findings table: The Severity column indicates the configured finding weight (CRITICAL / HIGH / MEDIUM / LOW). The Context column identifies whether the match occurred inside executable code, an inline comment, or a string literal — legacy comment-context matches receive a ×1.5 score weight. The ⚡ bolt icon marks clustered matches: three or more patterns within a 10-line window, each receiving an additional ×1.5 density multiplier under the legacy scoring model.
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | src/chatterbox/models/s3gen/decoder.py | 244 | Forward pass of the UNet1DConditional model. Args: x: (B, 80, T) mask (_type_) | STRING |
| HIGH | src/chatterbox/models/s3gen/xvector.py | 16 | Perform padding for the list of tensors. Args: xs (List): List of Tensors [(T_1, `*`), (T_2, `*`), ..., (T_ | STRING |
| HIGH | src/chatterbox/models/s3gen/utils/mask.py | 19 | def subsequent_mask( size: int, device: torch.device = torch.device("cpu"), ) -> torch.Tensor: """C | STRING |
| HIGH | src/chatterbox/models/s3gen/utils/mask.py | 60 | Create mask for subsequent steps (size, size) with chunk size, this is for streaming encoder Args: s | STRING |
| HIGH | src/chatterbox/models/s3gen/utils/mask.py | 168 | Make mask tensor containing indices of padded part. See description of make_non_pad_mask. Args: length | STRING |
| HIGH | src/chatterbox/models/s3gen/matcha/decoder.py | 364 | Forward pass of the UNet1DConditional model. Args: x (torch.Tensor): shape (batch_size, in_channels | STRING |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | src/chatterbox/tts.py | 0 | quick cleanup func for punctuation from llms or containing chars not seen often in the dataset | STRING |
| HIGH | src/chatterbox/tts_turbo.py | 0 | quick cleanup func for punctuation from llms or containing chars not seen often in the dataset | STRING |
| HIGH | src/chatterbox/mtl_tts.py | 0 | quick cleanup func for punctuation from llms or containing chars not seen often in the dataset | STRING |
| HIGH | src/chatterbox/tts.py | 0 | conditionals for t3 and s3gen - t3 conditionals: - speaker_emb - clap_emb - cond_prompt_speech_tokens - cond_prompt_spee | STRING |
| HIGH | src/chatterbox/tts_turbo.py | 0 | conditionals for t3 and s3gen - t3 conditionals: - speaker_emb - clap_emb - cond_prompt_speech_tokens - cond_prompt_spee | STRING |
| HIGH | src/chatterbox/mtl_tts.py | 0 | conditionals for t3 and s3gen - t3 conditionals: - speaker_emb - clap_emb - cond_prompt_speech_tokens - cond_prompt_spee | STRING |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | example_tts_turbo.py | 2 | CODE | |
| LOW | example_tts_nano.py | 2 | CODE | |
| LOW | src/chatterbox/__init__.py | 9 | CODE | |
| LOW | src/chatterbox/__init__.py | 10 | CODE | |
| LOW | src/chatterbox/__init__.py | 11 | CODE | |
| LOW | src/chatterbox/__init__.py | 11 | CODE | |
| LOW | src/chatterbox/models/tokenizers/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/tokenizers/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/s3tokenizer/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/s3tokenizer/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/s3tokenizer/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/s3tokenizer/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/s3tokenizer/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/t3/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/t3/t3.py | 4 | CODE | |
| LOW | src/chatterbox/models/t3/t3.py | 4 | CODE | |
| LOW | src/chatterbox/models/t3/inference/t3_hf_backend.py | 4 | CODE | |
| LOW | src/chatterbox/models/t3/modules/learned_pos_emb.py | 1 | CODE | |
| LOW | src/chatterbox/models/t3/modules/learned_pos_emb.py | 4 | CODE | |
| LOW | src/chatterbox/models/voice_encoder/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/voice_encoder/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/s3gen/__init__.py | 1 | CODE | |
| LOW | src/chatterbox/models/s3gen/__init__.py | 2 | CODE | |
| LOW | src/chatterbox/models/s3gen/flow.py | 23 | CODE | |
| LOW | src/chatterbox/models/s3gen/flow_matching.py | 14 | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | src/chatterbox/models/t3/t3.py | 281 | speech_head=self.speech_head, | COMMENT |
| LOW | src/chatterbox/models/s3gen/decoder.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Zhihao Du) | COMMENT |
| LOW | src/chatterbox/models/s3gen/flow.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Zhihao Du) | COMMENT |
| LOW | src/chatterbox/models/s3gen/hifigan.py | 1 | # jrm: adapted from CosyVoice/cosyvoice/hifigan/generator.py | COMMENT |
| LOW | src/chatterbox/models/s3gen/f0_predictor.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Kai Hu) | COMMENT |
| LOW | src/chatterbox/models/s3gen/s3gen.py | 1 | # Modified from CosyVoice https://github.com/FunAudioLLM/CosyVoice | COMMENT |
| LOW | src/chatterbox/models/s3gen/flow_matching.py | 1 | # Copyright (c) 2024 Alibaba Inc (authors: Xiang Lyu, Zhihao Du) | COMMENT |
| LOW | src/chatterbox/models/s3gen/transformer/attention.py | 1 | # Copyright (c) 2019 Shigeki Karita | COMMENT |
| LOW | src/chatterbox/models/s3gen/transformer/attention.py | 161 | Returns: | COMMENT |
| LOW | src/chatterbox/models/s3gen/transformer/attention.py | 281 | # cache(1, head, 0, d_k * 2) (16/-1, -1/-1, 16/0 mode) | COMMENT |
| LOW | src/chatterbox/models/s3gen/transformer/subsampling.py | 1 | # Copyright (c) 2021 Mobvoi Inc (Binbin Zhang, Di Wu) | COMMENT |
| LOW | src/chatterbox/models/s3gen/transformer/convolution.py | 1 | # Copyright (c) 2020 Mobvoi Inc. (authors: Binbin Zhang, Di Wu) | COMMENT |
| LOW | …hatterbox/models/s3gen/transformer/upsample_encoder.py | 1 | # Copyright (c) 2021 Mobvoi Inc (Binbin Zhang, Di Wu) | COMMENT |
| LOW | src/chatterbox/models/s3gen/transformer/embedding.py | 1 | # Copyright (c) 2020 Mobvoi Inc. (authors: Binbin Zhang, Di Wu) | COMMENT |
| LOW | …c/chatterbox/models/s3gen/transformer/encoder_layer.py | 1 | # Copyright (c) 2021 Mobvoi Inc (Binbin Zhang, Di Wu) | COMMENT |
| LOW | src/chatterbox/models/s3gen/transformer/activation.py | 1 | # Copyright (c) 2020 Johns Hopkins University (Shinji Watanabe) | COMMENT |
| LOW | …/models/s3gen/transformer/positionwise_feed_forward.py | 1 | # Copyright (c) 2019 Shigeki Karita | COMMENT |
| LOW | src/chatterbox/models/s3gen/utils/class_utils.py | 1 | # Copyright [2023-11-28] <sxc19@mails.tsinghua.edu.cn, Xingchen Song> | COMMENT |
| LOW | src/chatterbox/models/s3gen/utils/mask.py | 1 | # Copyright (c) 2019 Shigeki Karita | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | src/chatterbox/models/s3gen/flow_matching.py | 63 | CODE | |
| MEDIUM | src/chatterbox/models/s3gen/flow_matching.py | 64 | CODE | |
| MEDIUM | src/chatterbox/models/s3gen/flow_matching.py | 66 | CODE | |
| MEDIUM | src/chatterbox/models/s3gen/flow_matching.py | 69 | CODE | |
| MEDIUM | src/chatterbox/models/s3gen/flow_matching.py | 70 | CODE | |
| MEDIUM | src/chatterbox/models/s3gen/flow_matching.py | 71 | CODE | |
| MEDIUM | src/chatterbox/models/s3gen/flow_matching.py | 73 | CODE | |
| MEDIUM | src/chatterbox/models/s3gen/flow_matching.py | 74 | CODE | |
| MEDIUM | src/chatterbox/models/s3gen/flow_matching.py | 76 | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | src/chatterbox/tts_turbo.py | 272 | CODE | |
| LOW | src/chatterbox/models/t3/t3.py | 226 | CODE | |
| LOW | src/chatterbox/models/s3gen/decoder.py | 100 | CODE | |
| LOW | src/chatterbox/models/s3gen/flow.py | 44 | CODE | |
| LOW | src/chatterbox/models/s3gen/hifigan.py | 291 | CODE | |
| LOW | src/chatterbox/models/s3gen/xvector.py | 91 | CODE | |
| LOW | src/chatterbox/models/s3gen/xvector.py | 127 | CODE | |
| LOW | src/chatterbox/models/s3gen/xvector.py | 341 | CODE | |
| LOW | …hatterbox/models/s3gen/transformer/upsample_encoder.py | 101 | CODE | |
| LOW | src/chatterbox/models/s3gen/utils/mask.py | 86 | CODE | |
| LOW | src/chatterbox/models/s3gen/utils/mel.py | 21 | CODE | |
| LOW | src/chatterbox/models/s3gen/utils/mel.py | 36 | CODE | |
| LOW | src/chatterbox/models/s3gen/matcha/decoder.py | 61 | CODE | |
| LOW | src/chatterbox/models/s3gen/matcha/decoder.py | 162 | CODE | |
| LOW | src/chatterbox/models/s3gen/matcha/decoder.py | 201 | CODE | |
| LOW | src/chatterbox/models/s3gen/matcha/transformer.py | 159 | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| HIGH | gradio_tts_turbo_app.py | 62 | if (end < current_text.length && current_text[end] === ' ') suffix = ""; | CODE |
| HIGH | gradio_tts_app.py | 61 | min_p = gr.Slider(0.00, 1.00, step=0.01, label="min_p || Newer Sampler. Recommend 0.02 > 0.1. Handles Hi | CODE |
| HIGH | gradio_tts_app.py | 62 | top_p = gr.Slider(0.00, 1.00, step=0.01, label="top_p || Original Sampler. 1.0 Disables(recommended). Or | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW⚡ | multilingual_app.py | 150 | except Exception as e: | CODE |
| MEDIUM⚡ | multilingual_app.py | 151 | print(f"Error loading model: {e}") | CODE |
| LOW⚡ | multilingual_app.py | 158 | except Exception as e: | CODE |
| LOW | src/chatterbox/tts_turbo.py | 209 | except Exception as e: | CODE |
| LOW | src/chatterbox/tts_turbo.py | 236 | except Exception as e: | CODE |
| MEDIUM | src/chatterbox/tts_turbo.py | 228 | def norm_loudness(self, wav, sr, target_lufs=-27): | CODE |
| LOW | src/chatterbox/models/tokenizers/tokenizer.py | 131 | except Exception as e: | CODE |
| LOW | src/chatterbox/models/tokenizers/tokenizer.py | 187 | except Exception as e: | CODE |
| LOW | src/chatterbox/models/tokenizers/tokenizer.py | 251 | except Exception as e: | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| MEDIUM | README.md | 46 | | **Chatterbox-Multilingual V3** [(Language list)](#supported-languages) | **50 | CODE |
| MEDIUM | README.md | 48 | | Chatterbox [(Tips and Tricks)](#original-chatterbox-tips) | 500M | CODE |
| MEDIUM | src/chatterbox/models/s3gen/s3gen.py | 343 | # hallucination prevention, drop special tokens | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | src/chatterbox/tts_turbo.py | 24 | logger = logging.getLogger(__name__) | CODE |
| LOW | src/chatterbox/models/tokenizers/tokenizer.py | 18 | logger = logging.getLogger(__name__) | CODE |
| LOW | src/chatterbox/models/t3/t3.py | 6 | logger = logging.getLogger(__name__) | CODE |
| LOW | src/chatterbox/models/t3/t3.py | 30 | logger = logging.getLogger(__name__) | CODE |
| LOW | src/chatterbox/models/s3gen/flow.py | 18 | logger = logging.getLogger(__name__) | CODE |
| LOW | src/chatterbox/models/s3gen/flow.py | 27 | logger = logging.getLogger(__name__) | CODE |
| LOW | src/chatterbox/models/s3gen/utils/mel.py | 7 | logger = logging.getLogger(__name__) | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | src/chatterbox/models/tokenizers/tokenizer.py | 75 | CODE | |
| LOW | src/chatterbox/models/tokenizers/tokenizer.py | 285 | CODE | |
| LOW | src/chatterbox/models/s3gen/decoder.py | 229 | CODE | |
| LOW | src/chatterbox/models/s3gen/xvector.py | 130 | CODE | |
| LOW | src/chatterbox/models/s3gen/utils/mask.py | 89 | CODE | |
| LOW | src/chatterbox/models/s3gen/matcha/decoder.py | 345 | CODE | |
| LOW | src/chatterbox/models/s3gen/matcha/transformer.py | 96 | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | multilingual_app.py | 120 | def get_supported_languages_display() -> str: | CODE |
| LOW | src/chatterbox/models/t3/inference/t3_hf_backend.py | 34 | def prepare_inputs_for_generation( | CODE |
| LOW | src/chatterbox/models/t3/modules/perceiver.py | 22 | def _relative_position_bucket(relative_position, causal=True, num_buckets=32, max_distance=128): | CODE |
| LOW | src/chatterbox/models/t3/modules/perceiver.py | 84 | def scaled_dot_product_attention(self, q, k, v, mask=None): | CODE |
| LOW | src/chatterbox/models/s3gen/utils/mel.py | 15 | def dynamic_range_compression_torch(x, C=1, clip_val=1e-5): | CODE |
| LOW | src/chatterbox/models/s3gen/utils/intmeanflow.py | 5 | def get_intmeanflow_time_mixer(dims): | CODE |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | src/chatterbox/vc.py | 63 | # Check if MPS is available on macOS | COMMENT |
| LOW | src/chatterbox/tts.py | 169 | # Check if MPS is available on macOS | COMMENT |
| LOW | src/chatterbox/tts_turbo.py | 193 | # Check if MPS is available on macOS | COMMENT |
| LOW | src/chatterbox/mtl_tts.py | 233 | # Check if MPS is available on macOS | COMMENT |
| Severity | File | Line | Snippet | Context |
|---|---|---|---|---|
| LOW | src/chatterbox/models/s3gen/transformer/convolution.py | 124 | # It's better we just return None if no cache is required, | COMMENT |