# Scale Labs SEAL Korean (SEAL Korean)

> Scale's SEAL leaderboard for Korean: expert raters judge how well models answer about 1,000 prompts written by native Korean speakers. BenchLM shows it as display-only reference data, excluded from rankings. Scale marked this board deprecated as of March 2025 and no longer adds models.

Canonical page: https://benchlm.ai/benchmarks/scale-korean

- Category: [external](/external)
- Last updated: November 12, 2026 snapshot

## About SEAL Korean

- Year: 2026
- Tasks: About 1,000 prompts per language
- Format: Expert-rated score
- Difficulty: External agent and model evaluation
- Paper: [Scale Labs leaderboard](https://labs.scale.com/leaderboard/korean)

Single-turn prompts across general and culturally specific scenarios, rated by native-speaking experts. Scale publishes this board as a 0 to 100 score rather than a rating. BenchLM mirrors 19 published rows captured on November 12, 2026 snapshot.

SEAL Korean is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (19 models)

| Rank | Model | Creator | Score |
|------|-------|---------|-------|
| 1 | [o1-preview](/models/o1-preview) | OpenAI | 66.43 |
| 2 | [Claude 3.7 Sonnet (February 2025)](https://labs.scale.com/leaderboard/korean) | Anthropic | 64.93 |
| 3 | [GPT-4o (May 2024)](https://labs.scale.com/leaderboard/korean) | OpenAI | 64.58 |
| 4 | [GPT-4.5 Preview (February 2025)](https://labs.scale.com/leaderboard/korean) | OpenAI | 63.76 |
| 5 | [GPT-4 Turbo Preview](https://labs.scale.com/leaderboard/korean) | OpenAI | 60.76 |
| 6 | [Gemini 1.5 Pro (August 27, 2024)](https://labs.scale.com/leaderboard/korean) | Google | 60.28 |
| 7 | [GPT-4o (August 2024)](https://labs.scale.com/leaderboard/korean) | OpenAI | 59.93 |
| 8 | [Claude 3.5 Sonnet (June 2024)](https://labs.scale.com/leaderboard/korean) | Anthropic | 59.38 |
| 9 | [Gemini 2.0 Flash](/models/gemini-2-0-flash) | Google | 55.60 |
| 10 | [Claude 3 Sonnet](https://labs.scale.com/leaderboard/korean) | Anthropic | 54.17 |
| 11 | [Claude 3 Opus](/models/claude-3-opus) | Anthropic | 52.78 |
| 12 | [GPT-4o mini](/models/gpt-4o-mini) | OpenAI | 51.74 |
| 13 | [GPT-4](https://labs.scale.com/leaderboard/korean) | OpenAI | 51.39 |
| 14 | [Mistral Large 2](/models/mistral-large-2) | Mistral | 50.35 |
| 15 | [Llama 3.1 405B Instruct](https://labs.scale.com/leaderboard/korean) | Meta | 50.35 |
| 16 | [Gemini 1.5 Pro (May 2024)](https://labs.scale.com/leaderboard/korean) | Google | 40.42 |
| 17 | [Llama 3.1 70B Instruct](https://labs.scale.com/leaderboard/korean) | Meta | 37.23 |
| 18 | [Command R+](https://labs.scale.com/leaderboard/korean) | Cohere | 30.21 |
| 19 | [Llama 3.1 8B Instruct](https://labs.scale.com/leaderboard/korean) | Meta | 17.42 |

## FAQ

### What does SEAL Korean measure?

Scale's SEAL leaderboard for Korean: expert raters judge how well models answer about 1,000 prompts written by native Korean speakers. BenchLM shows it as display-only reference data, excluded from rankings. Scale marked this board deprecated as of March 2025 and no longer adds models.

### Which model leads the published SEAL Korean snapshot?

o1-preview currently leads the published SEAL Korean snapshot with a score of 66.43.

### How many models are evaluated on SEAL Korean?

The November 12, 2026 snapshot contains 19 AI models.

### Does SEAL Korean affect BenchLM's overall score?

Not directly. SEAL Korean is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.