All Benchmarks/Multilingual

Horangi 4 — GLP (general language)

KoreanHigher is better

Horangi 4's General Language Performance (GLP) average: a weighted average of 11 areas, including Korean grammar and meaning, expression, knowledge, and math and logical reasoning, plus coding and function calling. Those last two are English and code tasks, so it is not a pure measure of Korean ability. Scores run from 0 to 100%. Higher is better.

Other versions:HAE-RAE Bench v1 (Reading)Horangi 4 — ALT (alignment)Horangi 4 (Korean LLM Leaderboard)HRM8KIFEval-KoKMMLUKMMLU-ProKo-AIME 2025Ko-ARC-AGIKo-HLEKo-MoralKo-MTBenchKo-TruthfulQAKoBALT-700 (semantics)KoBALT-700 (syntax)KoBBQ (bias)Korean Hate Speech
Top score83.9%Claude Fable 5
Models tested49
Last updated2026.08.03
Model release date (newest on the right)
Best score so farHigher on the chart is better
  1. 1
    Anthropic
    Claude Fable 5
    2026.06.09 · high-effort
    83.9%
  2. 2
    Google
    Gemini 3.1 Pro
    2026.02.19 · high-effort
    83.4%
  3. 3
    OpenAI
    GPT-5.5
    2026.04.23 · xhigh-effort
    82.8%
  4. 4
    OpenAI
    GPT-5.6 Sol
    2026.07.09 · max-effort
    82.7%
  5. 5
    Google
    Gemini 3.5 Flash
    2026.05.19
    82.3%
  6. 6
    OpenAI
    GPT-5.4
    2026.03.05 · xhigh-effort
    81.9%
  7. 7
    OpenAI
    GPT-5.6 Terra
    2026.07.09 · max-effort
    81.6%
  8. 8
    xAI
    Grok 4.5
    2026.07.08
    80.2%
  9. 9
    Anthropic
    Claude Sonnet 5
    2026.06.30 · high-effort
    79.3%
  10. 10
    Alibaba
    Qwen3.7 Max
    2026.05.21
    78.8%

The "harness" is the agent program the AI used to carry out the task. The same model can score very differently depending on its harness and reasoning effort.