Sailor2
Open multilingual LLMs purpose-built for Southeast Asian languages
About
Sailor2 is Sea AI Lab's family of open multilingual models (1B, 8B, 20B) built for Southeast Asian languages, released February 2025 under Apache 2.0. Continued pre-training on 500B tokens covers 13+ SEA languages while keeping English and Chinese ability, and the team reports the 20B model matching GPT-4o head-to-head across SEA languages. It builds on Qwen2.5, so its context and tooling follow that lineage; the lowest-resource SEA languages still see uneven quality.
Specs
Context window
Pricing
Free quota
API access
Weights
Released
Benchmarks
Team-reported: Sailor2-20B achieves a 50-50 win rate against GPT-4o across SEA languages.
Provider-reported unless a source says otherwise. Benchmarks are marketing-adjacent by nature — treat them as directional.
Use cases
Trust score — 7/10
Scored 7/10 (verified tier, metrics-v1). Evidence: Paper arXiv:2502.12982 (2025-02-18) by Sea AI Lab and partners; Apache 2.0 confirmed on HF card and GitHub; 1B/8B/20B sizes; 500B continued-pretraining tokens over 13+ SEA languages; built on Qwen2.5; GPT-4o win-rate claim is team-reported and labeled as such; entry notes uneven quality on lowest-resource languages. License: Apache-2.0 confirmed on the Hugging Face card and GitHub repo..
How the trust score is built
Full methodology →- Adoption · stars/downloads, log-scaled vs peers
- 7.0/10
- Maintenance · recency, releases, issue load
- 7.0/10
- Openness · license + docs
- 10.0/10
- Manual review · sanity read + red-flag check
- 7.0/10
- · Paper arXiv:2502.12982 (2025-02-18) by Sea AI Lab and partners; Apache 2.0 confirmed on HF card and GitHub
- · 1B/8B/20B sizes; 500B continued-pretraining tokens over 13+ SEA languages; built on Qwen2.5
- · GPT-4o win-rate claim is team-reported and labeled as such; entry notes uneven quality on lowest-resource languages