reasoning benchmark
BFCL v2
Berkeley Function Calling Leaderboard (BFCL) v2 is a comprehensive benchmark for evaluating large language models' function calling capabilities. It features 2,251 question-function-answer pairs with enterprise and OSS-contributed functions, addressing data contamination and bias through live, user-contributed scenarios. The benchmark evaluates AST accuracy, executable accuracy, irrelevance detection, and relevance detection across multiple programming languages (Python, Java, JavaScript) and includes complex real-world function calling scenarios with multi-lingual prompts.
Updated Aug 7, 2026
Published models5
Registry coverage5
MetricScore
EvidenceB
BFCL v2 leaderboard
Sorted by the source-provided rank. Higher score is better according to the registry.
| 1 | ME | 77.3% | 100.0% | 5 | C | |
| 2 | NV | 74.1% | 75.0% | 5 | C | |
| 3 | NV | 73.7% | 50.0% | 5 | C | |
| 4 | ME | 67.0% | 25.0% | 5 | C | |
| 5 | NV | 63.6% | 0.0% | 5 | C |
Score distribution
Top published rows on the benchmark's original scale.
BFCL v2 highlights
The top published results on this benchmark's own scale.
What is BFCL v2?
Definition and scoring fields from the benchmark registry.
Berkeley Function Calling Leaderboard (BFCL) v2 is a comprehensive benchmark for evaluating large language models' function calling capabilities. It features 2,251 question-function-answer pairs with enterprise and OSS-contributed functions, addressing data contamination and bias through live, user-contributed scenarios. The benchmark evaluates AST accuracy, executable accuracy, irrelevance detection, and relevance detection across multiple programming languages (Python, Java, JavaScript) and includes complex real-world function calling scenarios with multi-lingual prompts.
Scores are shown in ratio. The current registry marks this benchmark as not independently verified with evidence level B.
- Family
- BFCL v2
- Modality
- text
- Primary category
- reasoning
- Score direction
- higher
- LLMBoard eligible
- No
- Evaluation key
- bfcl-v2|llm-stats-current
Source-native results are preserved. Eligibility for the overall LLMBoard score is a separate policy decision.
FAQ
Common questions about BFCL v2.
Which model scores highest on BFCL v2?
Llama 3.3 70B Instruct is currently ranked first with 77.3%.
What does BFCL v2 measure?
Berkeley Function Calling Leaderboard (BFCL) v2 is a comprehensive benchmark for evaluating large language models' function calling capabilities. It features 2,251 question-function-answer pairs with enterprise and OSS-contributed functions, addressing data contamination and bias through live, user-contributed scenarios. The benchmark evaluates AST accuracy, executable accuracy, irrelevance detection, and relevance detection across multiple programming languages (Python, Java, JavaScript) and includes complex real-world function calling scenarios with multi-lingual prompts.
Is a higher score better?
Yes. Higher values rank better for this benchmark.
How many models are compared?
5 unique published model results are currently shown.
Does this benchmark affect the overall score?
This benchmark is preserved as source-native evidence but is not eligible for the current overall score.