math benchmark
BBH Leaderboard
Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.
Updated Aug 17, 2026
Models12
Model coverage12
MetricScore
EvidenceB
BBH Ranking
Higher score ranks better on this benchmark.
Sort by
Rank | Model | Score | Percentile | Participants | Evidence | Evaluated |
|---|
| Rank01 | ModelAC | Score88.9% | Percentile100.0% | Participants12 | EvidenceC | Evaluated |
| Rank02 | ModelXI | Score88.4% | Percentile90.9% | Participants12 | EvidenceC | Evaluated |
| Rank03 | ModelAM Nova ProAmazon | Score86.9% | Percentile81.8% | Participants12 | EvidenceC | Evaluated |
| Rank04 | ModelAC | Score84.5% | Percentile72.7% | Participants12 | EvidenceC | Evaluated |
| Rank05 | ModelDE | Score84.3% | Percentile63.6% | Participants12 | EvidenceC | Evaluated |
| Rank06 | ModelAMNova LiteAmazon | Score82.4% | Percentile54.5% | Participants12 | EvidenceC | Evaluated |
| Rank07 | ModelAC | Score82.4% | Percentile45.5% | Participants12 | EvidenceC | Evaluated |
| Rank08 | ModelOP | Score81.5% | Percentile36.4% | Participants12 | EvidenceC | Evaluated |
| Rank09 | ModelAMNova MicroAmazon | Score79.5% | Percentile27.3% | Participants12 | EvidenceC | Evaluated |
| Rank10 | ModelAC | Score78.2% | Percentile18.2% | Participants12 | EvidenceC | Evaluated |
| Rank11 | ModelNR | Score67.8% | Percentile9.1% | Participants12 | EvidenceC | Evaluated |
| Rank12 | ModelBA | Score30.4% | Percentile0.0% | Participants12 | EvidenceC | Evaluated |
BBH Highlights
The leading models and scores on this benchmark.
BBH Score Distribution
A closer view of the leading scores on this benchmark.
The Top AI Models for BBH
The first five results on this benchmark, with official price and output speed added where the model identity can be matched.
Ranking basisThis bbh AI model leaderboard uses descending score in the benchmark's original unit. The leaderboard ranking keeps matched price and speed data separate from benchmark evidence.
- 01ACQwen3 235B A22BAlibaba Cloud / Qwen Team
- Score
- 88.9%
- Price
- $0.70 input / $2.8 output per 1M tokens
- Speed
- Up to 68 tok/s via Fireworks
Strengths
- Ranks #1 of 12 compared models
- 100th percentile on this benchmark
- C evidence result
Considerations
- This result measures BBH, not total model capability
- 02XIMiMo-V2.5-ProXiaomi
- Score
- 88.4%
- Price
- $0.43 input / $0.87 output per 1M tokens
- Speed
- Up to 69 tok/s via Novita
Strengths
- Ranks #2 of 12 compared models
- 91th percentile on this benchmark
- C evidence result
Considerations
- This result measures BBH, not total model capability
- 03AMNova ProAmazon
- Score
- 86.9%
- Price
- $0.80 input / $3.2 output per 1M tokens
- Speed
- Up to 100 tok/s via Bedrock
Strengths
- Ranks #3 of 12 compared models
- 82th percentile on this benchmark
- C evidence result
Considerations
- This result measures BBH, not total model capability
- 04ACQwen2.5 32B InstructAlibaba Cloud / Qwen Team
- Score
- 84.5%
- Price
- $0.70 input / $2.8 output per 1M tokens
Strengths
- Ranks #4 of 12 compared models
- 73th percentile on this benchmark
- C evidence result
Considerations
- This result measures BBH, not total model capability
- 05DEDeepSeek-V2.5DeepSeek
- Score
- 84.3%
- Speed
- Up to 100 tok/s via DeepSeek
Strengths
- Ranks #5 of 12 compared models
- 64th percentile on this benchmark
- C evidence result
Considerations
- This result measures BBH, not total model capability
Selection summary
Best AI Models for BBH
Qwen3 235B A22B currently leads BBH with 88.9%. It is the top model on this specific benchmark, while the best LLM for the broader task should also be checked against other benchmarks, price and runtime.
Use this leaderboard with the supporting benchmark results and coverage details above. A leaderboard position summarizes the selected ranking signal; it does not replace workload-specific testing.
What is BBH?
What BBH measures and how its scores work.
Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.
Scores are shown in ratio. This benchmark is not independently verified and has an evidence level of B.
- Family
- BBH
- Modality
- text
- Primary category
- math
- Score direction
- higher
- LLMBoard eligible
- Yes
- Evaluation key
- bbh|llm-stats-current
Benchmark scores retain their original unit. Overall score eligibility is shown separately.
FAQ
Common questions about BBH.
Which model scores highest on BBH?
Qwen3 235B A22B is currently ranked first with 88.9%.
What does BBH measure?
Big-Bench Hard (BBH) is a suite of 23 challenging tasks selected from BIG-Bench for which prior language model evaluations did not outperform the average human-rater. These tasks require multi-step reasoning across diverse domains including arithmetic, logical reasoning, reading comprehension, and commonsense reasoning. The benchmark was designed to test capabilities believed to be beyond current language models and focuses on evaluating complex reasoning skills including temporal understanding, spatial reasoning, causal understanding, and deductive logical reasoning.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
12 model results are currently shown.
Does this benchmark affect the overall score?
Yes. This benchmark can contribute to the current LLMBoard capability score.