BFCL-V4

BFCL-V4 - Berkeley Function Calling Leaderboard version 4, evaluating function calling and tool use capabilities.

Category

General Agents

Max Score

—

Score Type

percent

Active

Yes

Model Scores

Scores for BFCL-V4

Model Score Date Verified
Qwen3.5-4B
47.86%
01.02.2026 Verified
Qwen3.5-122B-A10B
100.00%
01.02.2026 Verified
Qwen3.5-27B
91.19%
01.02.2026 Unverified
Qwen3.5-35B-A3B
88.33%
01.02.2026 Unverified
Qwen3-235B-A22B
58.57%
01.02.2026 Unverified
GPT-5-mini 2025-08-07
60.24%
01.02.2026 Unverified
Spark-X2.5-4B
83.10%
— Verified
Spark-X2.5-1.7B
39.76%
— Verified
Qwen3.5-9B
85.48%
— Verified
Qwen3.5-2B
31.90%
— Verified
Gemma4-12B
17.14%
— Verified
Gemma4-E4B
15.95%
— Verified
Gemma4-E2B
30.20
— Verified
MiniCPM5-2B
86.67%
— Verified
LFM2.5-2.6B
73.57%
— Verified
granite-4.2-3B
52.38%
— Verified
Nemotron-3-Nano-4B
32.14%
— Verified
LFM2.5-8B-A1B
45.24%
— Verified