Benchmarks / Berkeley Function Calling Leaderboard (BFCL) V4

Reported by Berkeley Function Calling Leaderboard (BFCL) V4

Berkeley Function Calling Leaderboard (BFCL) V4

Completing tasks over several turns with stateful tools, including when functions or parameters are missing.

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: Completing tasks over several turns with stateful tools, including when functions or parameters are missing.

Latest results. , 109 configurations ranked.

Top 3 on Multi-turn tasks (% correct, higher is better):

  1. xLAM-2-70b-fc-r (Salesforce): 77.4%
  2. xLAM-2-8b-fc-r (Salesforce): 70.0%
  3. xLAM-2-32b-fc-r (Salesforce): 69.5%

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.

Multi-turn tasks: xLAM-2-32b-fc-r

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

  1. 1 xLAM-2-70b-fc-rFunction callingSalesforce 77.4%
  2. 2 xLAM-2-8b-fc-rFunction callingSalesforce 70.0%
  3. 3 xLAM-2-32b-fc-rFunction callingSalesforce 69.5%
  4. 4 Claude Opus 4.5Function callingAnthropic 68.4%
  5. 5 GLM-4.6 (reasoning on)Function callingZ.ai 68.0%
  6. 6 Gemini 3 Pro PreviewFunction callingGoogle 63.1%
  7. 7 BitAgent Bounty 8BMode not statedBittensor 62.4%
  8. 8 o3PromptingOpenAI 62.3%
  9. 9 Claude Sonnet 4.5Function callingAnthropic 61.4%
  10. 10 Gemini 3 Pro PreviewPromptingGoogle 60.8%
  11. 11 Grok 4.1 Fast (reasoning)Function callingxAI 58.9%
  12. 12 xLAM-2-3b-fc-rFunction callingSalesforce 58.4%
  13. 13 Arch-Agent-32BMode not statedKatanemo 54.3%
  14. 14 Claude Haiku 4.5Function callingAnthropic 53.6%
  15. 15 Nanbeige4-3B-Thinking (2511)Function callingNanbeige 51.1%

Full results

BFCL: multi-turn tasks, % correct, higher is better
#ModelMulti-turn tasks
% correct, higher is better
Price
$ per million tokens, in / out
1 xLAM-2-70b-fc-rFunction callingSalesforce
77.4%
–
2 xLAM-2-8b-fc-rFunction callingSalesforce
70.0%
–
3 xLAM-2-32b-fc-rFunction callingSalesforce
69.5%
–
4 Claude Opus 4.5Function callingAnthropic
68.4%
$5 / $25
5 GLM-4.6 (reasoning on)Function callingZ.ai
68.0%
$0.50 / $2
6 Gemini 3 Pro PreviewFunction callingGoogle
63.1%
–
7 BitAgent Bounty 8BMode not statedBittensor
62.4%
–
8 o3PromptingOpenAI
62.3%
$2 / $8
9 Claude Sonnet 4.5Function callingAnthropic
61.4%
$3 / $15
10 Gemini 3 Pro PreviewPromptingGoogle
60.8%
–
11 Grok 4.1 Fast (reasoning)Function callingxAI
58.9%
–
12 xLAM-2-3b-fc-rFunction callingSalesforce
58.4%
–
13 Arch-Agent-32BMode not statedKatanemo
54.3%
–
14 Claude Haiku 4.5Function callingAnthropic
53.6%
$1 / $5
15 Nanbeige4-3B-Thinking (2511)Function callingNanbeige
51.1%
–
16 Kimi K2 (0711)Function callingMoonshot AI
50.6%
$0.57 / $2.30
17 Command A ReasoningFunction callingCohere
50.1%
–
18 Qwen3-32BFunction callingAlibaba
47.9%
$0.14 / $0.40
19 Grok 4 (0709)PromptingxAI
47.0%
–
20 Grok 4.1 Fast (non-reasoning)Function callingxAI
46.8%
–
21 Qwen3-235B-A22B-Instruct-2507Function callingAlibaba
45.4%
$0.15 / $0.75
22 DeepSeek-V3.2-Exp (reasoning on)PromptingDeepSeek
44.9%
$0.27 / $0.41
23 Qwen3-235B-A22B-Instruct-2507PromptingAlibaba
44.6%
$0.15 / $0.75
24 GPT-5.2PromptingOpenAI
43.8%
$1.75 / $14
25 Qwen3-32BPromptingAlibaba
43.3%
$0.14 / $0.40
26 Qwen3-8BFunction callingAlibaba
41.8%
–
26 o4-miniFunction callingOpenAI
41.8%
–
28 Nanbeige3.5-Pro-ThinkingFunction callingNanbeige
40.0%
–
29 GPT-4.1Function callingOpenAI
38.9%
$2 / $8
30 ToolACE-2-8BFunction callingHuawei Noah's Ark Lab and USTC
38.4%
–
31 DeepSeek-V3.2-ExpFunction callingDeepSeek
37.4%
$0.27 / $0.41
32 Gemini 2.5 FlashFunction callingGoogle
36.3%
$0.30 / $2.50
33 xLAM-2-1b-fc-rFunction callingSalesforce
36.0%
–
34 Arch-Agent-3BMode not statedKatanemo
34.9%
–
35 Qwen3-14BFunction callingAlibaba
34.8%
$0.10 / $0.22
36 GPT-5 nanoFunction callingOpenAI
34.5%
$0.05 / $0.40
37 GPT-4.1 miniFunction callingOpenAI
34.1%
$0.40 / $1.60
38 Grok 4 (0709)Function callingxAI
33.9%
–
39 Qwen3-8BPromptingAlibaba
33.4%
–
40 Qwen3-30B-A3B-Instruct-2507Function callingAlibaba
30.0%
$0.09 / $0.30
41 Command AFunction callingCohere
29.5%
$2.50 / $10
42 GPT-5.2Function callingOpenAI
28.1%
$1.75 / $14
43 GPT-5 miniFunction callingOpenAI
27.5%
$0.25 / $2
44 Arch-Agent-1.5BMode not statedKatanemo
26.6%
–
45 Qwen3-14BPromptingAlibaba
26.1%
$0.10 / $0.22
46 Hammer2.1-7BFunction callingMadeAgents
23.9%
–
47 GPT-4.1 nanoFunction callingOpenAI
23.6%
$0.10 / $0.40
48 Qwen3-30B-A3B-Instruct-2507PromptingAlibaba
23.5%
$0.09 / $0.30
49 Qwen3-4B-Instruct-2507Function callingAlibaba
22.1%
–
50 Llama 3.3 70B InstructFunction callingMeta
21.5%
$0.71 / $0.71
51 Qwen3-4B-Instruct-2507PromptingAlibaba
20.5%
–
52 Llama 4 Maverick 17B 128E Instruct (FP8)Function callingMeta
20.3%
–
53 Gemini 2.5 FlashPromptingGoogle
16.8%
$0.30 / $2.50
54 o4-miniPromptingOpenAI
16.6%
–
55 Hammer2.1-3BFunction callingMadeAgents
16.5%
–
56 Claude Opus 4.5PromptingAnthropic
16.1%
$5 / $25
57 Hammer2.1-1.5BFunction callingMadeAgents
15.6%
–
58 Mistral Small 3.2 24BPromptingMistral AI
14.8%
$0.094 / $0.25
58 o3Function callingOpenAI
14.8%
$2 / $8
60 Mistral Large 2.1 (2411)Function callingMistral AI
14.1%
–
61 Mistral Large 2.1 (2411)PromptingMistral AI
13.8%
–
62 Gemini 2.5 Flash-LiteFunction callingGoogle
13.5%
$0.10 / $0.40
63 Mistral Small 3.2 24BFunction callingMistral AI
11.5%
$0.094 / $0.25
64 Llama 3.1 8B InstructPromptingMeta
11.1%
$0.05 / $0.08
65 Qwen3-1.7BFunction callingAlibaba
11.0%
–
66 Gemma 3 27BPromptingGoogle
10.8%
$0.08 / $0.35
66 Mistral Medium 3Function callingMistral AI
10.8%
$0.40 / $2
68 CoALM 70BMode not statedUIUC and Oumi
10.6%
–
69 Mistral Medium 3Mode not statedMistral AI
9.9%
$0.40 / $2
70 GPT-4.1PromptingOpenAI
9.8%
$2 / $8
71 Llama 4 ScoutFunction callingMeta
9.0%
$0.10 / $0.30
72 Command R7B (12-2024)Function callingCohere
8.3%
$0.037 / $0.15
73 CoALM 8BMode not statedUIUC and Oumi
8.0%
–
74 Mistral NemoFunction callingMistral AI
7.8%
$0.029 / $0.03
75 Gemini 2.5 Flash-LitePromptingGoogle
7.6%
$0.10 / $0.40
76 Granite 3.1 8B InstructFunction callingIBM
7.5%
–
77 Granite 3.2 8B InstructFunction callingIBM
7.4%
–
78 Falcon3 10B InstructFunction callingTII
6.5%
–
79 Gemma 3 12BPromptingGoogle
5.8%
$0.05 / $0.15
80 GPT-5 miniPromptingOpenAI
5.5%
$0.25 / $2
81 Granite 20B FunctionCallingFunction callingIBM
5.4%
–
82 Falcon3 7B InstructFunction callingTII
5.0%
–
83 Llama 3.2 3B InstructFunction callingMeta
4.0%
$0.05 / $0.33
84 Phi-4PromptingMicrosoft
3.9%
$0.07 / $0.14
84 MiniCPM3-4BFunction callingOpenBMB
3.9%
–
86 Qwen3-0.6BFunction callingAlibaba
3.6%
–
87 MiniCPM3-4BPromptingOpenBMB
3.5%
–
88 Hammer2.1-0.5BFunction callingMadeAgents
2.9%
–
88 RZN-TPromptingPhronetic AI
2.9%
–
90 Bielik 11B v2.3 InstructPromptingSpeakLeash and ACK Cyfronet AGH
2.6%
–
91 Granite 4.0 350MFunction callingIBM
2.5%
–
91 GPT-4.1 miniPromptingOpenAI
2.5%
$0.40 / $1.60
93 Nova 2 LiteFunction callingAmazon
2.1%
$0.30 / $2.50
94 GPT-4.1 nanoPromptingOpenAI
2.0%
$0.10 / $0.40
95 Nova Pro 1.0Function callingAmazon
1.9%
$0.80 / $3.20
96 Claude Haiku 4.5PromptingAnthropic
1.8%
$1 / $5
97 Claude Sonnet 4.5PromptingAnthropic
1.6%
$3 / $15
98 Qwen3-0.6BPromptingAlibaba
1.4%
–
98 Nova Micro 1.0Function callingAmazon
1.4%
$0.035 / $0.14
100 Falcon3 3B InstructFunction callingTII
1.0%
–
101 Mistral NemoPromptingMistral AI
0.8%
$0.029 / $0.03
101 GPT-5 nanoPromptingOpenAI
0.8%
$0.05 / $0.40
103 Gemma 3 4BPromptingGoogle
0.4%
$0.04 / $0.05
103 Palmyra X 004Function callingWriter
0.4%
–
105 Gemma 3 1BPromptingGoogle
0.0%
–
105 Llama 3.2 1B InstructFunction callingMeta
0.0%
$0.027 / $0.20
105 Ministral 8B (2410)Function callingMistral AI
0.0%
–
105 Llama 3.1 Nemotron Ultra 253B v1Function callingNVIDIA
0.0%
–
105 Falcon3 1B InstructFunction callingTII
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

Multi-turn tasks, 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.

-20%0%20%40%60%80% Jul 20242025Jul2026Jul Mistral NemoLlama 3.1 8B InstructMistral Large 2.1 (2411)Llama 3.3 70B InstructGPT-4.1o3GLM-4.6Claude Opus 4.5

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.

AnthropicZ.aiGoogleOpenAIxAIMoonshot AI

0%20%40%60%80% Jul 20242025Jul2026Jul AnthropicZ.aiGoogleOpenAIxAIMoonshot AI

What it measures

Completing tasks over several turns with stateful tools, including when functions or parameters are missing.

What it does not measure

Not open-ended agent work: the tools and tasks are BFCL's simulated APIs.

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