Aleph Alpha's open-weight English-German mixture-of-experts model for self-hosting, released under Apache 2.0 in October 2026
Index-Translate-35B-A3B (preview)
New· 4Open-weight translation model from Bilibili's Index team, covering 150 languages and following glossary and style instructions
About Index-Translate-35B-A3B (preview)
The Index team at Bilibili, the Chinese video platform, has released Index-Translate-35B-A3B in preview. It is an open-weight translation model, and its Hugging Face repository dates from 2 October 2026. The weights are free to download under the Apache 2.0 license.
Model design and size
The architecture is a mixture-of-experts design resting on the Qwen 3.5 base. Its total size is 35 billion parameters, and about 3 billion are active on each token. The family also includes smaller dense models with 9 billion and 2 billion parameters. The model card says it translates text across 150 languages.
Instruction-following translation
Beyond plain translation, the model can take direction on terms, layout, tone, structure, context and output length. A user can ask it to keep to a glossary or move a text from a formal to a casual tone. Those controls matter to businesses that need consistent product names and a consistent voice across many languages. The developers' technical report compares it with other translation models on standard benchmarks.
Running it on your own infrastructure
Because the model is open weight, it can be run on a company's own servers rather than sent to a third-party service. The license permits commercial use, which is the main appeal for businesses that want translation kept inside their own infrastructure.
TrustList summary
Index-Translate-35B-A3B (preview) is an open-weight translation model from Bilibili's Index group, which belongs to the Chinese video platform. It was posted on Hugging Face on 2 October 2026 under Apache 2.0. Businesses that want translation on their own infrastructure are the natural audience.
What Index-Translate-35B-A3B (preview) offers
- Translation: text translated across 150 languages.
- Instruction following: follows directions on wording rules, layout, tone, structure, context and output length, such as holding to a term list or moving from formal to casual.
- Model family: a mixture-of-experts design of 35 billion parameters, about 3 billion active for each token, plus smaller dense 9B and 2B siblings.
- Benchmarks: results come from the developer's own technical report, which compares the model with other translation models.
Who it works with
The model appeals to businesses that want translation hosted on their own servers, and its licence permits commercial use.
Key facts
- Published: 2 October 2026 on Hugging Face
- Licence: Apache 2.0
- Base model: Qwen 3.5
- Parameters: 35 billion in total, about 3 billion active
Summary written by TrustList from the business's own published information, Nov 2026.
Benchmarks & AI stats
| Benchmark | Official | Community avg |
|---|---|---|
| No benchmark scores yet. Be the first to add one. | ||
“Official” values are editor-approved and feed the ranking. “Community avg” is the mean of member submissions (shown for transparency; it never affects the ranking until an editor approves a value).
Sign in to add a benchmark score for this model.
Index-Translate-35B-A3B (preview) launches
Categories & features
Trust Score
An earned signal from verification, reviews, awards, transparency and engagement - the vendor can't buy it.
Updated 10/4/2026
Request a demo or quote from Index-Translate-35B-A3B (preview)
Reviews
Write the first review of Index-Translate-35B-A3B (preview)
Used it? Your experience helps other buyers decide.
Write a reviewQuestions & answers
No questions yet. Be the first to ask about Index-Translate-35B-A3B (preview).
Index-Translate-35B-A3B (preview) alternatives
Similar listings buyers compare against this one.
Diffusion language model from Inception for enterprise teams building real-time voice agents, announced in September 2026
Open-weight agentic coding model from IQuestLab with a 524,288-token context, for teams that self-host
OpenAI's mid-tier GPT-6 model with a 1.05-million token context, image input and multi-agent support in beta
Perplexity's open-weight decision model that picks among given options for routing and checks in AI systems
Mid-tier Anthropic model for everyday work, coding and document creation, with a 1 million token context window
Google's frontier model, announced in September 2026 and released first to selected cyber defenders
Cantina Security's first open-weight security model, a post-trained GLM-5.3-Flash for vulnerability research and exploit checks
More on TrustList
Everything here links back to the same verified catalogue. Pick your next stop.
- Write a reviewHelp other buyers decide
- CompaniesAgencies, consultancies and IT service providers, ranked by verified reviews.
- ProductsSoftware and SaaS with pricing, features, integrations and alternatives.
- ArticlesGuides, comparisons and research from the editorial desk and the community.
- AwardsAnnual recognition decided by verified reviews and an independent jury.
- LaunchesNew products and releases, voted up by the community every day.
- RequestsBuyers describe what they need; vendors respond directly.
- PeopleReviewers, authors and makers with public profiles.
- ComparePut up to four listings side by side before you shortlist.