BFCL v3 Leaderboard

Berkeley Function Calling Leaderboard v3 — testing function/tool calling accuracy.

Data from LayerLens

As of November 21, 2026, the top-scoring model on BFCL v3 is GLM 4.5 at 76.7%, followed by Claude Opus 4.7 at 76.6% and Gemini 3.1 Flash Lite Preview at 76.5%. 20 models have been evaluated on this benchmark.

Last updated: November 21, 2026

8 Ways to Use Fewer Tokens

Models

20

Best Score

76.7

Average

61.6

Std Dev

15.2

Categories
Multi-turn
Sponsored The Grid | Spot Priced LLM API
The Grid | Spot Priced LLM API — Why pay list price when providers will bid for your API usage? Up to 80% off.
View spot prices
Provider
Model
Input $/M
Output $/M
BFCL v3
Actions
$0.600
$2.200
76.7
$5.000
$25.000
76.6
$0.250
$1.500
76.5
$0.064
$0.224
75.7
$0.064
$0.224
75.7
$0.780
$3.900
74.9
$0.060
$0.400
74.6
$0.060
$0.400
74.6
$0.130
$0.850
69.1
$0.800
$3.200
67.9
$0.490
$2.500
64.5
$0.490
$2.500
64.5
$0.100
$0.300
55.7
$0.250
$0.750
54.3
$0.500
$3.000
53.5
$0.000
$0.000
47.8
$0.060
$0.240
41.6
$0.060
$0.240
41.6
$0.070
$0.140
40.8
$0.550
$2.200
25.3

Pricing from OpenRouter. Benchmarks from Artificial Analysis.

About BFCL v3

Berkeley Function Calling Leaderboard v3 — testing function/tool calling accuracy.

This leaderboard shows all models with BFCL v3 benchmark scores, ranked from highest to lowest. Pricing data is included to help you compare performance against cost.

Frequently Asked Questions

Berkeley Function Calling Leaderboard v3 — testing function/tool calling accuracy.
As of November 21, 2026, GLM 4.5 leads the BFCL v3 leaderboard with a score of 76.7. Rankings change as new models are released and evaluated.
Currently 20 models have been evaluated on BFCL v3, with an average score of 61.6 and standard deviation of 15.2.
Benchmark scores are updated when new evaluations are published by our data sources (Artificial Analysis and LayerLens). Pricing data is refreshed daily from OpenRouter.