Benchmarks / Berkeley Function Calling Leaderboard (BFCL) V4

Reported by Berkeley Function Calling Leaderboard (BFCL) V4

Berkeley Function Calling Leaderboard (BFCL) V4

Choosing the right function and arguments in one turn, on BFCL's curated set.

Data older than 90 days: last updated 16 Dec 2025

Results dated
16 Dec 2025
Results
109 configurations of 81 models
Unit
% correct
Licence
Apache License 2.0

Berkeley Function Calling Leaderboard (BFCL) V4 in brief

What it measures. Berkeley Function Calling Leaderboard (BFCL) V4 is a third-party benchmark, shown here with credit. Measured here: Choosing the right function and arguments in one turn, on BFCL's curated set.

Latest results. , 109 configurations ranked.

Top 3 on Single-turn calls (curated) (% correct, higher is better):

  1. Gemini 3 Pro Preview (Google): 90.7%
  2. Qwen3-235B-A22B-Instruct-2507 (Alibaba): 90.3%
  3. Qwen3-32B (Alibaba): 90.3%

How to cite. Berkeley Function Calling Leaderboard (BFCL) V4, via Spring Prompt, results dated 16 December 2025. https://springprompt.com/benchmarks/bfcl These results belong to Berkeley Function Calling Leaderboard (BFCL) V4, not to us. Licence: Apache License 2.0. Credit Berkeley Function Calling Leaderboard (BFCL) V4 first.

Single-turn calls (curated): Qwen3-235B-A22B-Instruct-2507

Top 15 of 109 results · % correct, higher is better. Choose a model to highlight it.Clear highlight

  1. 1 Gemini 3 Pro PreviewPromptingGoogle 90.7%
  2. 2 Qwen3-235B-A22B-Instruct-2507PromptingAlibaba 90.3%
  3. 2 Qwen3-32BPromptingAlibaba 90.3%
  4. 4 Mistral Small 3.2 24BPromptingMistral AI 89.7%
  5. 4 Claude Opus 4.5PromptingAnthropic 89.7%
  6. 6 xLAM-2-32b-fc-rFunction callingSalesforce 89.6%
  7. 7 Qwen3-14BPromptingAlibaba 89.5%
  8. 8 Llama 4 ScoutFunction callingMeta 89.4%
  9. 9 Qwen3-30B-A3B-Instruct-2507PromptingAlibaba 88.9%
  10. 9 Arch-Agent-32BMode not statedKatanemo 88.9%
  11. 11 Qwen3-32BFunction callingAlibaba 88.8%
  12. 12 GPT-4.1PromptingOpenAI 88.7%
  13. 12 Claude Sonnet 4.5Function callingAnthropic 88.7%
  14. 12 Llama 4 Maverick 17B 128E Instruct (FP8)Function callingMeta 88.7%
  15. 15 Claude Opus 4.5Function callingAnthropic 88.6%
  16. 104 Qwen3-235B-A22B-Instruct-2507Function callingAlibaba 37.4%

Full results

BFCL: single-turn calls (curated), % correct, higher is better
#ModelSingle-turn calls (curated)
% correct, higher is better
Price
$ per million tokens, in / out
1 Gemini 3 Pro PreviewPromptingGoogle
90.7%
–
2 Qwen3-235B-A22B-Instruct-2507PromptingAlibaba
90.3%
$0.15 / $0.75
2 Qwen3-32BPromptingAlibaba
90.3%
$0.14 / $0.40
4 Mistral Small 3.2 24BPromptingMistral AI
89.7%
$0.094 / $0.25
4 Claude Opus 4.5PromptingAnthropic
89.7%
$5 / $25
6 xLAM-2-32b-fc-rFunction callingSalesforce
89.6%
–
7 Qwen3-14BPromptingAlibaba
89.5%
$0.10 / $0.22
8 Llama 4 ScoutFunction callingMeta
89.4%
$0.10 / $0.30
9 Qwen3-30B-A3B-Instruct-2507PromptingAlibaba
88.9%
$0.09 / $0.30
9 Arch-Agent-32BMode not statedKatanemo
88.9%
–
11 Qwen3-32BFunction callingAlibaba
88.8%
$0.14 / $0.40
12 GPT-4.1PromptingOpenAI
88.7%
$2 / $8
12 Claude Sonnet 4.5Function callingAnthropic
88.7%
$3 / $15
12 Llama 4 Maverick 17B 128E Instruct (FP8)Function callingMeta
88.7%
–
15 Claude Opus 4.5Function callingAnthropic
88.6%
$5 / $25
15 Qwen3-8BPromptingAlibaba
88.6%
–
17 Mistral NemoPromptingMistral AI
88.5%
$0.029 / $0.03
18 xLAM-2-70b-fc-rFunction callingSalesforce
88.4%
–
19 Grok 4.1 Fast (reasoning)Function callingxAI
88.3%
–
20 Grok 4.1 Fast (non-reasoning)Function callingxAI
88.1%
–
20 Gemini 2.5 FlashPromptingGoogle
88.1%
$0.30 / $2.50
22 Llama 3.3 70B InstructFunction callingMeta
88.0%
$0.71 / $0.71
23 Qwen3-4B-Instruct-2507Function callingAlibaba
87.9%
–
24 Qwen3-8BFunction callingAlibaba
87.6%
–
24 Command AFunction callingCohere
87.6%
$2.50 / $10
24 GLM-4.6 (reasoning on)Function callingZ.ai
87.6%
$0.50 / $2
27 Palmyra X 004Function callingWriter
87.5%
–
28 Gemma 3 27BPromptingGoogle
87.2%
$0.08 / $0.35
29 ToolACE-2-8BFunction callingHuawei Noah's Ark Lab and USTC
87.1%
–
30 Nova 2 LiteFunction callingAmazon
87.0%
$0.30 / $2.50
31 Arch-Agent-3BMode not statedKatanemo
86.7%
–
32 Gemini 2.5 Flash-LiteFunction callingGoogle
86.6%
$0.10 / $0.40
32 Nova Pro 1.0Function callingAmazon
86.6%
$0.80 / $3.20
34 Claude Haiku 4.5Function callingAnthropic
86.5%
$1 / $5
35 Qwen3-4B-Instruct-2507PromptingAlibaba
86.4%
–
36 Command A ReasoningFunction callingCohere
86.3%
–
37 Qwen3-30B-A3B-Instruct-2507Function callingAlibaba
85.8%
$0.09 / $0.30
37 Gemini 3 Pro PreviewFunction callingGoogle
85.8%
–
39 DeepSeek-V3.2-Exp (reasoning on)PromptingDeepSeek
85.5%
$0.27 / $0.41
39 Hammer2.1-7BFunction callingMadeAgents
85.5%
–
41 Grok 4 (0709)Function callingxAI
85.4%
–
42 Mistral Medium 3Mode not statedMistral AI
85.3%
$0.40 / $2
43 Falcon3 10B InstructFunction callingTII
85.0%
–
43 Gemini 2.5 FlashFunction callingGoogle
85.0%
$0.30 / $2.50
43 Hammer2.1-3BFunction callingMadeAgents
85.0%
–
46 Qwen3-14BFunction callingAlibaba
84.9%
$0.10 / $0.22
46 CoALM 8BMode not statedUIUC and Oumi
84.9%
–
48 Mistral Large 2.1 (2411)Function callingMistral AI
84.7%
–
49 GPT-4.1 miniPromptingOpenAI
84.6%
$0.40 / $1.60
49 xLAM-2-8b-fc-rFunction callingSalesforce
84.6%
–
51 Llama 3.1 8B InstructPromptingMeta
84.0%
$0.05 / $0.08
52 Gemini 2.5 Flash-LitePromptingGoogle
83.9%
$0.10 / $0.40
53 GPT-4.1 miniFunction callingOpenAI
83.8%
$0.40 / $1.60
54 CoALM 70BMode not statedUIUC and Oumi
83.4%
–
55 Mistral Large 2.1 (2411)PromptingMistral AI
83.0%
–
55 Hammer2.1-1.5BFunction callingMadeAgents
83.0%
–
55 xLAM-2-3b-fc-rFunction callingSalesforce
83.0%
–
58 Qwen3-1.7BFunction callingAlibaba
82.9%
–
59 Mistral NemoFunction callingMistral AI
82.8%
$0.029 / $0.03
59 GPT-4.1Function callingOpenAI
82.8%
$2 / $8
59 Grok 4 (0709)PromptingxAI
82.8%
–
62 Falcon3 7B InstructFunction callingTII
82.7%
–
62 Arch-Agent-1.5BMode not statedKatanemo
82.7%
–
62 Llama 3.2 3B InstructFunction callingMeta
82.7%
$0.05 / $0.33
65 Granite 20B FunctionCallingFunction callingIBM
82.4%
–
66 o3PromptingOpenAI
81.9%
$2 / $8
66 GPT-5.2Function callingOpenAI
81.9%
$1.75 / $14
68 MiniCPM3-4BFunction callingOpenBMB
81.8%
–
69 BitAgent Bounty 8BMode not statedBittensor
81.6%
–
69 Kimi K2 (0711)Function callingMoonshot AI
81.6%
$0.57 / $2.30
69 Nanbeige4-3B-Thinking (2511)Function callingNanbeige
81.6%
–
72 Bielik 11B v2.3 InstructPromptingSpeakLeash and ACK Cyfronet AGH
81.5%
–
73 o4-miniPromptingOpenAI
81.3%
–
74 Command R7B (12-2024)Function callingCohere
81.0%
$0.037 / $0.15
75 GPT-5 nanoPromptingOpenAI
80.8%
$0.05 / $0.40
76 Granite 3.2 8B InstructFunction callingIBM
79.8%
–
77 Gemma 3 12BPromptingGoogle
79.4%
$0.05 / $0.15
78 Granite 3.1 8B InstructFunction callingIBM
78.3%
–
78 GPT-5.2PromptingOpenAI
78.3%
$1.75 / $14
80 Nova Micro 1.0Function callingAmazon
74.1%
$0.035 / $0.14
81 Mistral Small 3.2 24BFunction callingMistral AI
73.6%
$0.094 / $0.25
82 GPT-4.1 nanoFunction callingOpenAI
73.0%
$0.10 / $0.40
83 GPT-4.1 nanoPromptingOpenAI
72.4%
$0.10 / $0.40
84 Qwen3-0.6BFunction callingAlibaba
71.8%
–
85 MiniCPM3-4BPromptingOpenBMB
70.5%
–
86 Qwen3-0.6BPromptingAlibaba
70.0%
–
87 GPT-5 miniFunction callingOpenAI
69.9%
$0.25 / $2
88 Phi-4PromptingMicrosoft
69.6%
$0.07 / $0.14
89 xLAM-2-1b-fc-rFunction callingSalesforce
69.0%
–
90 GPT-5 miniPromptingOpenAI
68.0%
$0.25 / $2
90 GPT-5 nanoFunction callingOpenAI
68.0%
$0.05 / $0.40
92 RZN-TPromptingPhronetic AI
67.9%
–
92 Granite 4.0 350MFunction callingIBM
67.9%
–
94 Mistral Medium 3Function callingMistral AI
67.4%
$0.40 / $2
95 Hammer2.1-0.5BFunction callingMadeAgents
66.0%
–
96 Gemma 3 4BPromptingGoogle
61.1%
$0.04 / $0.05
97 Claude Sonnet 4.5PromptingAnthropic
59.8%
$3 / $15
98 Claude Haiku 4.5PromptingAnthropic
55.4%
$1 / $5
99 Falcon3 3B InstructFunction callingTII
54.6%
–
100 o3Function callingOpenAI
40.4%
$2 / $8
101 Llama 3.2 1B InstructFunction callingMeta
38.4%
$0.027 / $0.20
101 Nanbeige3.5-Pro-ThinkingFunction callingNanbeige
38.4%
–
103 o4-miniFunction callingOpenAI
37.7%
–
104 Qwen3-235B-A22B-Instruct-2507Function callingAlibaba
37.4%
$0.15 / $0.75
105 DeepSeek-V3.2-ExpFunction callingDeepSeek
34.9%
$0.27 / $0.41
106 Gemma 3 1BPromptingGoogle
20.2%
–
107 Falcon3 1B InstructFunction callingTII
9.0%
–
108 Ministral 8B (2410)Function callingMistral AI
0.0%
–
108 Llama 3.1 Nemotron Ultra 253B v1Function callingNVIDIA
0.0%
–

Swipe the table sideways for more columns.

Source: Berkeley Function Calling Leaderboard (BFCL) V4, via Spring Prompt, springprompt.com/benchmarks/bfcl. Test your own prompt on these models →

Ranks follow the score as shown, so equal numbers share a rank. Results as published by Berkeley Function Calling Leaderboard (BFCL) V4; we do not re-run them.

Progress over time

Single-turn calls (curated), plotted at each model's release date, higher is better. Each model counts once, at its best setting. Scores are today's, not launch day's; 38 models without a known release date are left off. Hover or tap a mark for its model.

Best score so far

Every model at its release date. The line moves each time a new release beats every earlier one, and each new best is named where there is room.

0%50%100% Jul 20242025Jul2026Jul Mistral NemoQwen3-32BQwen3-235B-A22B-Instruct-2507Gemini 3 Pro Preview

Best score by lab

Each lab's best so far, stepping whenever one of its releases beats its own earlier models. The 6 labs with the best scores today; 7 others not drawn.

GoogleAlibabaMistral AIAnthropicMetaOpenAI

82%84%86%88%90%92% Jul 20242025Jul2026Jul GoogleAlibabaMistral AIAnthropicMetaOpenAI

What it measures

Choosing the right function and arguments in one turn, on BFCL's curated set.

What it does not measure

Not reliability on your own tools: BFCL's functions and queries are a fixed test set, and a correct call is judged by its form, not by what it achieved.

Berkeley Function Calling Leaderboard (BFCL) V4 by the UC Berkeley Gorilla team, Apache License 2.0.