Vendor · Language models · Audio models
Bilibili Index Team has 11 models in the directory: 7 language models and 4 audio models.
In the directory: 7 language models and 4 audio models. 11 of them with open weights that can run locally.
Every model there ever was. Models with a profile lead to every benchmark, price and source.
| Index-Translate-2B | Sep 30, 2026 | Open weights | n/a | n/a | n/a | n/a | 37 |
| Index-Translate-35B-A3B Preview | Sep 30, 2026 | Preview | n/a | n/a | n/a | n/a | 32 |
| Index-Translate-9B | Sep 30, 2026 | Open weights | n/a | n/a | n/a | n/a | 37 |
| Index-Homura-2B | n/a | Open weights | n/a | n/a | n/a | n/a | 23 |
| Index-Homura-9B | n/a | Open weights | n/a | n/a | n/a | n/a | 23 |
| Index-NativeLong-2B | n/a | Open weights | n/a | n/a | n/a | n/a | 23 |
| Index-NativeLong-9B | n/a | Open weights | n/a | n/a | n/a | n/a | 24 |
| Index-Echo-S2ST-2B | n/a | Open weights | n/a | 30 |
| Index-Echo-S2ST-9B | n/a | Open weights | n/a | 29 |
| Index-Echo-S2TT-2B | n/a | Open weights | n/a | 26 |
| Index-Echo-S2TT-9B | n/a | Open weights | n/a | 24 |
Language model · Sep 30, 2026
Index-Translate-2B is built for text translation with explicit constraints on terminology, style, and output format. Its model card lists 150 languages and also presents the translation of community expressions as a use case.
Language model · Sep 30, 2026
Index-Translate-9B is the larger variant of Index-Translate-2B for text translation across 150 languages. Translation instructions can specify terminology, style, and output format; translating community expressions is also a documented use case.
Audio model · n/a
Index-Echo-S2ST-2B adds a spoken target version to subtitle translation. An S2TT backbone, a Hidden2CV mapper, and CosyVoice3 form its pipeline. The input voice serves as a reference, though Bilibili notes that similarity can vary.
Audio model · n/a
Index-Echo-S2ST-9B combines the larger S2TT text decoder with Hidden2CV and CosyVoice3 for speech output. It translates Chinese to English, Japanese, or Spanish, and English to Chinese, Japanese, or Spanish. The input voice guides synthesis without guaranteeing an identical copy.
Audio model · n/a
Index-Echo-S2TT-2B translates speech into subtitles. Its package uses a 2B text decoder, accepts audio or video, and returns a transcript plus translated, timestamped text segments for Chinese to English, Japanese, or Spanish.
Audio model · n/a
Index-Echo-S2TT-9B uses a larger text decoder than the 2B variant, while retaining the same subtitle output. The source transcript and translation appear as timestamped text segments for Chinese to English, Japanese, or Spanish.
Language model · n/a
Index-Homura-2B accepts a target syllable count for its translation. The model adjusts the wording toward that target, while a glossary can further guide word choice. The count remains approximate and does not guarantee a particular spoken duration.
Language model · n/a
Index-Homura-9B offers the same syllable control as the 2B variant in a larger checkpoint. Its translation is designed to preserve meaning while approaching a requested syllable count. That target is approximate and does not guarantee the duration of the spoken line.
Language model · n/a
Index-NativeLong-2B translates a complete document in one generation so information from distant passages remains available. Its published templates cover Chinese to and from English, and Chinese to and from Japanese.
Language model · n/a
Index-NativeLong-9B offers the document translation of the 2B variant in a larger checkpoint. It translates the whole document in one step to account for terms and references across distant passages. Its four templates connect Chinese in both directions with English or Japanese.
| Vendor | Models | Model types | Best language model | Top 10 | Open weights | New in 2026 |
|---|---|---|---|---|---|---|
| Anthropic | 33 | 33 language models | Claude Opus 5.5 (Rank 1 to 3) | 4 | 0 | 13 |
| OpenAI | 63 | 46 language models, 4 image models, 2 video models, 9 audio models, 2 specialist models | GPT-6 Astra (Rank 4 to 8) | 2 | 4 | 27 |
| 66 | 43 language models, 4 image models, 6 video models, 11 audio models, 2 specialist models | Gemini 4 Carbon (Rank 1 to 5) | 2 | 19 | 26 | |
| Meta | 32 | 29 language models, 1 image model, 1 video model, 1 specialist model | Muse Wonder Full (Rank 6 to 9) | 2 | 26 | 7 |
| DeepSeek | 23 | 23 language models | DeepSeek V4.1 Flash 1017 (Rank 13 to 20) | 0 | 23 | 9 |
| Moonshot AI | 12 | 12 language models | Kimi K3.1 Pro (Rank 10 to 13) | 0 | 10 | 7 |
| xAI | 16 | 12 language models, 2 image models, 2 video models | Grok 4.7 (Rank 18 to 24) | 0 | 1 | 12 |
| Bilibili Index Team | 11 | 7 language models, 4 audio models | not ranked yet | 0 | 11 | 3 |
By release date. Grey: retired models.
3 releases since January 2026, by release date.
The directory lists 11 Bilibili Index models from Bilibili Index Team: 7 language models and 4 audio models. It counts every model Bilibili Index Team ever released, including retired ones. 10 are currently available, the newest being Index-Translate-2B, Index-Translate-9B and Index-Echo-S2ST-2B.
Index-Translate-2B, released on Sep 30, 2026. The timeline below shows every release by date.
Yes, 11 of the 11 models have open weights. Each model profile lists how much memory it needs.
Data as of Nov 1, 2026