> For the complete documentation index, see [llms.txt](https://docs.sea-lion.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://docs.sea-lion.ai/models/sea-lion-v5.md).

# SEA-LION v5 (latest)

SEA-LION version 5, released in November 2026, scales our Southeast Asian multimodal and agentic capabilities from resource-constrained edge devices to massive enterprise reasoning clusters. Built on Gemma-4, Qwen3.5/Qwen4Exp, and Nemotron 3.5 Bolt foundations—and aligned via On-Policy Distillation (OPD)—this multi-architecture suite leverages expansive tokenizers supporting up to 8,900 native non-Latin characters to deliver state-of-the-art regional fluency and computational efficiency.

## Qwen-SEA-LION-v5 (397B, 27B & 9B Series)

Our flagship causal and vision-language series anchored by a 397B Mixture of Experts (MoE) model adapted to the Qwen4Exp architecture. It pairs QSA Hybrid Attention and Gated Residuals with a CPU-offloaded 328 GB N-gram Embedding memory table, delivering frontier multimodal reasoning and long-context efficiency across enterprise and mid-tier deployments.

* [Qwen-SEA-LION-v5-397B-A17B-IT](https://docs.sea-lion.ai/models/sea-lion-v5/qwen-sea-lion-v5.md)
* [Qwen-SEA-LION-v5-27B-IT](https://docs.sea-lion.ai/models/sea-lion-v5/qwen-sea-lion-v5.md)
* [Qwen-SEA-LION-v5-9B-IT](https://docs.sea-lion.ai/models/sea-lion-v5/qwen-sea-lion-v5.md)

  Refer to the detailed page to explore full benchmarks, architectural deep dives, and configuration details for the entire Qwen family:

  * Multimodal & Mixture-of-Experts (MoE) Foundations
  * CPU-Offloaded N-gram Memory & QSA Hybrid Attention Configurations
  * Lossless FP8 & NVFP4 Quantized Checkpoints

## Gemma-SEA-LION-v5 (E2B and E4B Series)

Our ultra-efficient edge series built on Gemma-4. Leveraging native support, it is optimized for low-latency multimodal processing, precise function-calling, and autonomous agentic tool-use on constrained devices. Nemotron 3.5 Bolt engineered for rapid inference speed and balanced reasoning upon Gemma4-E4B, it provides a compact, enterprise-ready footprint for regional text generation and structured workflows.

* [Gemma-SEA-LION-v5-E2B-IT and Nemotron-SEA-LION-v5-Bolt-E4B-IT](https://docs.sea-lion.ai/models/sea-lion-v5/gemma-sea-lion-v5.md)

  Refer to the detailed page to access model cards, usage scripts, and deployment configurations for the Gemma family:

  * Core Multimodal & Agentic Foundations for Gemma
  * Finetuned from Gemma4 for Nemotron 3.5 Bolt E4B


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://docs.sea-lion.ai/models/sea-lion-v5.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
