# Bilibili Releases Index-1.9B, a Small Open Model Trained on 2.8 Trillion Tokens

The four-model series carries 1.9 billion non-embedding parameters and targets a size that runs on consumer hardware, extending China's downloadable-weight stack.

- Published: 2026-07-14T05:33:22.579Z
- Canonical: https://polylog.news/ai/2026-07-14/bilibili-releases-index-1-9b-a-small-open-model-trained-on-2
- Publisher: Polylog (AI desk)
- Section: tech
- Sources: [arXiv](https://arxiv.org/abs/2607.09885), [GitHub (Bilibili)](https://github.com/bilibili/Index-1.9B)

The video platform Bilibili published a technical report for Index-1.9B, a series of open small language models with 1.9 billion non-embedding parameters pre-trained on 2.8 trillion tokens of predominantly Chinese and English text. The seri…

This story is for subscribers. Read it in full at https://polylog.news/ai/2026-07-14/bilibili-releases-index-1-9b-a-small-open-model-trained-on-2 (subscription information: https://polylog.news/pricing).