benchmark evidence
BBH (Open LLM Leaderboard)
Open LLM Leaderboard v2 — BBH (Open LLM Leaderboard).
activeupstream source ->
winner on BBH (Open LLM Leaderboard)
direct benchmark result, not a broad vertical composite | source row dated 2026-05-08
scored on 2026-05-08
latest mapped results | top 20
| # | Model | Score | Evidence | Tested |
|---|---|---|---|---|
| 1 | Meta: Llama 3.3 70B Instruct | 69.2 | model-only leaderboard | 2026-05-08 |
| 2 | Mistral Large | 67.0 | model-only leaderboard | 2026-05-08 |
| 3 | Mistral: Mistral Small 4 | 58.1 | model-only leaderboard | 2026-05-08 |
| 4 | Mistral: Mistral Nemo | 50.1 | model-only leaderboard | 2026-05-08 |
what this result means
Open LLM Leaderboard v2 — BBH (Open LLM Leaderboard).
This benchmark contributes direct public evidence. Read its scope before generalizing the result.
A win here is a win on BBH (Open LLM Leaderboard). Broad task pages require independent corroboration before naming a general winner.
source record
category: reasoning
metric: acc_norm
matched models: 4
latest source date: 2026-05-08
direction: higher is better