Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Loading page…
Preferred languages
More languages supported by Google Translate
Open the selector to load the current language catalog.
Translations are provided by Google Translate.
Agentic benchmark
Berkeley Function Calling Leaderboard v3. Every model the catalog carries a published BFCL v3 value for, ranked by that value.
A function-calling benchmark for tool selection, schema adherence, and argument correctness, covering single-turn, parallel, irrelevance and multi-turn subsets. 9 models in the ModelScale catalog have a published value.
Each point is a model with a published BFCL v3 value and complete token prices, on the 75% input / 25% output blend. Frontier models are spaced evenly along the cost axis.
9 models with a published BFCL v3 value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.
| Rank | Model | Provider | Tool invocation and schema evaluation |
|---|---|---|---|
| 1 | PMTernary Bonsai 2 27B | Prism ML | 74.9 |
| 2 | Alibaba | 70.8 | |
| 3 | Alibaba | 70.3 |
What the benchmark measures and how its value is published, from the source’s own record.
BFCL v3 (Berkeley Function Calling Leaderboard v3) is an agentic benchmark published by The Berkeley Function Calling Leaderboard (BFCL): From Tool Use to Agentic Evaluation of Large Language Models. A function-calling benchmark for tool selection, schema adherence, and argument correctness, covering single-turn, parallel, irrelevance and multi-turn subsets. Its evaluation set is described by the source as function-calling tasks. The source rates its difficulty as advanced tool use.
Each model's result is published by The Berkeley Function Calling Leaderboard (BFCL): From Tool Use to Agentic Evaluation of Large Language Models as tool invocation and schema evaluation. ModelScale prints that value to 1 decimal place, exactly as published, and never converts or rescales it. This page ranks the 9 models that carry a published value, highest value first; a model the source never scored is left out rather than ranked at zero. The source does not state whether a higher value is the better result, so neither does this page.
| 4 | Alibaba | 68.7 |
| 5 | NVIDIA | 68.5 |
| 6 | Alibaba | 68.0 |
| 7 | NVIDIA | 67.3 |
| 8 | Alibaba | 66.4 |
| 9 | Alibaba | 63.0 |
Evidence key: Observed
Rows are ordered by the value The Berkeley Function Calling Leaderboard (BFCL): From Tool Use to Agentic Evaluation of Large Language Models published, highest first. The source does not state whether a higher value is the better result, so this page does not either: for a benchmark that measures a rate of failure, read the table from the bottom.