bilibili
company · ChinaScores
1 product on the map — 1 open-ish.
Openness
3 high confidence- weights
- open (IndexTTS-2.5 checkpoints on the Hub, ungated)
- data
- closed (no training-data statement on the card or the README)
- code
- partial (inference code and a web UI; no training pipeline)
- license
- bilibili-Model-Use-License (IndexTTS-2.5 weights and the repository code; a separate license is required above 100 million monthly active users or RMB 1 billion annual revenue)
- superseded-release
- IndexTTS-2 (2025, the model the bilibili license was written for; IndexTTS-1.5 before it is tagged Apache-2.0 on the Hub)
IndexTTS-2.5 downloads freely, and the bilibili Model Use License allows commercial use for everyone except the very largest companies, which must ask for a separate license. Bilibili publishes inference code only and says nothing about the training data.
- https://huggingface.co/api/models?author=IndexTeam&limit=100 recorded 2026-11-12
Hub listing: IndexTTS-2.5 created 2026-08-10 tagged "license:other"; IndexTTS-2 created 2025-06-18 with no license tag; IndexTTS-1.5 created 2025-05-13 tagged "license:apache-2.0".
- https://huggingface.co/api/models/IndexTeam/IndexTTS-2.5 recorded 2026-11-12
Hub record: "gated":false, "license_name":"bilibili-model-license", with gpt.pth, s2mel.pth and codec.pth in the file list.
- https://huggingface.co/IndexTeam/IndexTTS-2.5/raw/main/LICENSE recorded 2026-11-12
License file in the IndexTTS-2.5 repository: "bilibili Model Use License Agreement"; 2.2: if your products "had more than 100 million monthly active users" or your annual revenue "exceeded RMB 1 billion, You must request a separated license from us"; 2.1 grants a "royalty-free limited license to Use the Model".
- https://huggingface.co/IndexTeam/IndexTTS-2.5/raw/main/README.md recorded 2026-11-12
Card: "**License:** [bilibili Model Use License Agreement](LICENSE)"; "Parameters: ~0.8B (GPT backbone)"; "Languages: Chinese, English, Japanese, Spanish, Arabic"; Getting Started covers download and inference only.
- https://raw.githubusercontent.com/index-tts/index-tts/HEAD/LICENSE recorded 2026-11-12
Repository LICENSE: the same "bilibili Model Use License Agreement", whose "Model" includes "model weights and final code ... published by us at https://github.com/index-tts/index-tts".
- https://raw.githubusercontent.com/index-tts/index-tts/HEAD/README.md recorded 2026-11-12
README: "`2026/08/10` We release **IndexTTS-2.5**"; installation, inference, web UI and evaluation sections and "This project is released under the [bilibili Model Use License Agreement](LICENSE)"; no training instructions.
Adoption
2 high confidenceHugging Face downloads over the trailing month, summed across the IndexTTS-2.5, IndexTTS-2 and IndexTTS-1.5 checkpoints. The ModelScope copies are not counted.
- https://huggingface.co/api/models?author=IndexTeam&limit=100 recorded 2026-11-12
"downloads":29259 for IndexTeam/IndexTTS-2.5, 22823 for IndexTeam/IndexTTS-2 and 2541 for IndexTeam/IndexTTS-1.5.
Capability
3 low confidenceIndexTTS has no listener rating yet, so it is placed on its own results on the public CV3-Eval test set, where its figures sit close to CosyVoice 3, OmniVoice and VoxCPM2.
- https://artificialanalysis.ai/text-to-speech/leaderboard recorded 2026-11-12
The 104-row leaderboard has no IndexTTS entry.
- https://raw.githubusercontent.com/index-tts/index-tts/HEAD/README.md recorded 2026-11-12
README "Table 1: Zero-shot TTS on CV3-Eval (Arabic uses an in-house test set)": rows for VoxCPM2, OmniVoice, Moss-TTS 1.5, CosyVoice3-0.5B, CosyVoice3-1.5B, FireRedTTS-2, Fish Audio S2 Pro, Qwen3-TTS, IndexTTS2.5 (Avg 6.75 / 73.18) and IndexTTS2.5-RL (Avg 6.00 / 73.63).