Apertus-SEA-LION-v4-8B-IT
Apertus-SEA-LION-v4-8B-IT adapts Apertus-8B-Instruct-2509 for Southeast Asian text tasks. Its post-training uses about 6.4 million instruction-text pairs. AI Singapore also publishes the post-training datasets and evaluation resources.
How this release works
The training mixture includes tool-calling examples. The card lists Indonesian, Vietnamese, Thai, Filipino, Tamil, Burmese and Malay as regional languages. The released training and evaluation material makes this a distinct option within the wider set of SEA-LION architectures.Apertus-SEA-LION-v4-8B-IT official model card
How to assess it for your work
Practical guidance. Consider Apertus when you need to inspect the material behind a regional adaptation. Trace an example from the intended task through the prompt, response and evaluation rule. Published resources can support that review, but they do not replace checking the model on the documents and wording your users will supply.
Specifications & access
| Detail | Value and conditions |
|---|---|
| Access | Downloadable |
| Model identifiers | aisingapore/Apertus-SEA-LION-v4-8B-IT |
| Model size | 8 billion parametersApertus-SEA-LION-v4-8B-IT official model card |
| Context stated by the card | 65K tokensApertus-SEA-LION-v4-8B-IT official model card |
| License declared by the model card | MITApertus-SEA-LION-v4-8B-IT official model card |
| Input and output | TextApertus-SEA-LION-v4-8B-IT official model card |
Related models
- Gemma-SEA-LION-v4-4B-VL
AI Singapore · SEA-LION v4
A 4B Gemma 3 vision-language model adapted for regional image understanding, visual parsing and tool calls.
Launch date not confirmedDownloadable - Qwen-SEA-LION-v4-32B-IT
AI Singapore · SEA-LION v4
A 32B Qwen3 text model with continued pretraining on Southeast Asian languages and regional instruction tuning.
Launch date not confirmedDownloadable