Benchmarks / Berkeley Function Calling Leaderboard (BFCL) V4

Reported by Berkeley Function Calling Leaderboard (BFCL) V4

Berkeley Function Calling Leaderboard (BFCL) V4

Storing and recalling facts across sessions with memory tools.

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: Storing and recalling facts across sessions with memory tools.

Latest results. , 109 configurations ranked.

Top 3 on Memory (% correct, higher is better):

  1. Claude Opus 4.5 (Anthropic): 73.8%
  2. Claude Sonnet 4.5 (Anthropic): 65.0%
  3. Gemini 3 Pro Preview (Google): 61.7%

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.

Memory: MiniCPM3-4B

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

  1. 1 Claude Opus 4.5Function callingAnthropic 73.8%
  2. 2 Claude Sonnet 4.5Function callingAnthropic 65.0%
  3. 3 Gemini 3 Pro PreviewPromptingGoogle 61.7%
  4. 4 Grok 4 (0709)Function callingxAI 55.9%
  5. 5 GLM-4.6 (reasoning on)Function callingZ.ai 55.7%
  6. 6 Gemini 3 Pro PreviewFunction callingGoogle 54.8%
  7. 7 Claude Haiku 4.5Function callingAnthropic 54.4%
  8. 8 DeepSeek-V3.2-ExpFunction callingDeepSeek 54.2%
  9. 9 Grok 4.1 Fast (reasoning)Function callingxAI 54.0%
  10. 10 o3PromptingOpenAI 51.8%
  11. 11 Grok 4 (0709)PromptingxAI 50.5%
  12. 12 o3Function callingOpenAI 47.3%
  13. 13 GPT-5.2Function callingOpenAI 45.8%
  14. 14 Nanbeige3.5-Pro-ThinkingFunction callingNanbeige 45.2%
  15. 15 GPT-5 miniFunction callingOpenAI 44.3%
  16. 67 MiniCPM3-4BFunction callingOpenBMB 12.0%
  17. 74 MiniCPM3-4BPromptingOpenBMB 9.5%

Full results

BFCL: memory, % correct, higher is better
#ModelMemory
% correct, higher is better
Price
$ per million tokens, in / out
1 Claude Opus 4.5Function callingAnthropic
73.8%
$5 / $25
2 Claude Sonnet 4.5Function callingAnthropic
65.0%
$3 / $15
3 Gemini 3 Pro PreviewPromptingGoogle
61.7%
–
4 Grok 4 (0709)Function callingxAI
55.9%
–
5 GLM-4.6 (reasoning on)Function callingZ.ai
55.7%
$0.50 / $2
6 Gemini 3 Pro PreviewFunction callingGoogle
54.8%
–
7 Claude Haiku 4.5Function callingAnthropic
54.4%
$1 / $5
8 DeepSeek-V3.2-ExpFunction callingDeepSeek
54.2%
$0.27 / $0.41
9 Grok 4.1 Fast (reasoning)Function callingxAI
54.0%
–
10 o3PromptingOpenAI
51.8%
$2 / $8
11 Grok 4 (0709)PromptingxAI
50.5%
–
12 o3Function callingOpenAI
47.3%
$2 / $8
13 GPT-5.2Function callingOpenAI
45.8%
$1.75 / $14
14 Nanbeige3.5-Pro-ThinkingFunction callingNanbeige
45.2%
–
15 GPT-5 miniFunction callingOpenAI
44.3%
$0.25 / $2
16 DeepSeek-V3.2-Exp (reasoning on)PromptingDeepSeek
44.1%
$0.27 / $0.41
17 Gemini 2.5 FlashFunction callingGoogle
41.3%
$0.30 / $2.50
18 Gemini 2.5 FlashPromptingGoogle
38.7%
$0.30 / $2.50
19 Nanbeige4-3B-Thinking (2511)Function callingNanbeige
36.8%
–
20 o4-miniPromptingOpenAI
35.3%
–
21 o4-miniFunction callingOpenAI
34.2%
–
22 GPT-5 miniPromptingOpenAI
29.3%
$0.25 / $2
23 Kimi K2 (0711)Function callingMoonshot AI
29.0%
$0.57 / $2.30
24 Command A ReasoningFunction callingCohere
28.8%
–
25 Gemma 3 12BPromptingGoogle
27.5%
$0.05 / $0.15
25 Falcon3 10B InstructFunction callingTII
27.5%
–
27 GPT-4.1 miniFunction callingOpenAI
26.9%
$0.40 / $1.60
28 Qwen3-32BFunction callingAlibaba
26.7%
$0.14 / $0.40
29 Grok 4.1 Fast (non-reasoning)Function callingxAI
26.2%
–
30 Mistral Large 2.1 (2411)Function callingMistral AI
25.0%
–
31 Phi-4PromptingMicrosoft
24.7%
$0.07 / $0.14
31 GPT-5 nanoFunction callingOpenAI
24.7%
$0.05 / $0.40
33 GPT-5 nanoPromptingOpenAI
24.5%
$0.05 / $0.40
34 GPT-4.1 miniPromptingOpenAI
24.3%
$0.40 / $1.60
35 Qwen3-235B-A22B-Instruct-2507Function callingAlibaba
23.9%
$0.15 / $0.75
35 Qwen3-4B-Instruct-2507PromptingAlibaba
23.9%
–
35 GPT-4.1Function callingOpenAI
23.9%
$2 / $8
38 Mistral Large 2.1 (2411)PromptingMistral AI
23.7%
–
39 Mistral Medium 3Function callingMistral AI
23.0%
$0.40 / $2
40 Mistral Medium 3Mode not statedMistral AI
21.7%
$0.40 / $2
41 GPT-4.1PromptingOpenAI
21.5%
$2 / $8
42 xLAM-2-32b-fc-rFunction callingSalesforce
20.9%
–
43 Gemini 2.5 Flash-LiteFunction callingGoogle
20.7%
$0.10 / $0.40
43 Falcon3 7B InstructFunction callingTII
20.7%
–
45 Qwen3-14BFunction callingAlibaba
19.6%
$0.10 / $0.22
46 Qwen3-235B-A22B-Instruct-2507PromptingAlibaba
19.4%
$0.15 / $0.75
47 Llama 4 Maverick 17B 128E Instruct (FP8)Function callingMeta
18.9%
–
47 GPT-4.1 nanoFunction callingOpenAI
18.9%
$0.10 / $0.40
49 ToolACE-2-8BFunction callingHuawei Noah's Ark Lab and USTC
18.5%
–
50 Mistral Small 3.2 24BFunction callingMistral AI
18.1%
$0.094 / $0.25
51 Qwen3-30B-A3B-Instruct-2507Function callingAlibaba
17.6%
$0.09 / $0.30
51 Qwen3-4B-Instruct-2507Function callingAlibaba
17.6%
–
53 GPT-4.1 nanoPromptingOpenAI
16.8%
$0.10 / $0.40
54 Command AFunction callingCohere
16.6%
$2.50 / $10
55 Qwen3-32BPromptingAlibaba
15.7%
$0.14 / $0.40
56 Mistral Small 3.2 24BPromptingMistral AI
15.1%
$0.094 / $0.25
57 Qwen3-8BFunction callingAlibaba
14.6%
–
57 Arch-Agent-32BMode not statedKatanemo
14.6%
–
59 Granite 3.1 8B InstructFunction callingIBM
14.4%
–
59 xLAM-2-70b-fc-rFunction callingSalesforce
14.4%
–
61 xLAM-2-8b-fc-rFunction callingSalesforce
14.0%
–
62 Gemma 3 27BPromptingGoogle
13.6%
$0.08 / $0.35
63 Qwen3-8BPromptingAlibaba
13.1%
–
63 Palmyra X 004Function callingWriter
13.1%
–
65 Gemini 2.5 Flash-LitePromptingGoogle
12.7%
$0.10 / $0.40
66 Granite 3.2 8B InstructFunction callingIBM
12.5%
–
67 MiniCPM3-4BFunction callingOpenBMB
12.0%
–
68 xLAM-2-3b-fc-rFunction callingSalesforce
11.4%
–
68 Bielik 11B v2.3 InstructPromptingSpeakLeash and ACK Cyfronet AGH
11.4%
–
70 Qwen3-14BPromptingAlibaba
11.2%
$0.10 / $0.22
71 Llama 3.1 8B InstructPromptingMeta
10.8%
$0.05 / $0.08
72 Mistral NemoFunction callingMistral AI
10.3%
$0.029 / $0.03
73 Qwen3-30B-A3B-Instruct-2507PromptingAlibaba
9.7%
$0.09 / $0.30
74 MiniCPM3-4BPromptingOpenBMB
9.5%
–
75 Qwen3-0.6BFunction callingAlibaba
8.6%
–
75 Gemma 3 4BPromptingGoogle
8.6%
$0.04 / $0.05
75 Mistral NemoPromptingMistral AI
8.6%
$0.029 / $0.03
78 Qwen3-0.6BPromptingAlibaba
8.4%
–
79 Arch-Agent-1.5BMode not statedKatanemo
8.2%
–
79 Llama 3.3 70B InstructFunction callingMeta
8.2%
$0.71 / $0.71
79 Llama 4 ScoutFunction callingMeta
8.2%
$0.10 / $0.30
82 Falcon3 3B InstructFunction callingTII
7.7%
–
83 Arch-Agent-3BMode not statedKatanemo
6.9%
–
83 RZN-TPromptingPhronetic AI
6.9%
–
85 Llama 3.2 3B InstructFunction callingMeta
6.2%
$0.05 / $0.33
86 Qwen3-1.7BFunction callingAlibaba
6.0%
–
87 Falcon3 1B InstructFunction callingTII
5.8%
–
87 CoALM 70BMode not statedUIUC and Oumi
5.8%
–
89 Claude Sonnet 4.5PromptingAnthropic
5.4%
$3 / $15
90 Command R7B (12-2024)Function callingCohere
5.2%
$0.037 / $0.15
91 Ministral 8B (2410)Function callingMistral AI
4.5%
–
92 GPT-5.2PromptingOpenAI
3.9%
$1.75 / $14
92 xLAM-2-1b-fc-rFunction callingSalesforce
3.9%
–
94 Gemma 3 1BPromptingGoogle
3.2%
–
94 Granite 4.0 350MFunction callingIBM
3.2%
–
94 Llama 3.2 1B InstructFunction callingMeta
3.2%
$0.027 / $0.20
97 Hammer2.1-3BFunction callingMadeAgents
3.0%
–
98 CoALM 8BMode not statedUIUC and Oumi
2.8%
–
99 Claude Haiku 4.5PromptingAnthropic
2.6%
$1 / $5
100 Nova 2 LiteFunction callingAmazon
2.4%
$0.30 / $2.50
100 Nova Micro 1.0Function callingAmazon
2.4%
$0.035 / $0.14
102 Nova Pro 1.0Function callingAmazon
1.9%
$0.80 / $3.20
102 Claude Opus 4.5PromptingAnthropic
1.9%
$5 / $25
104 BitAgent Bounty 8BMode not statedBittensor
1.5%
–
105 Hammer2.1-0.5BFunction callingMadeAgents
1.1%
–
106 Granite 20B FunctionCallingFunction callingIBM
0.0%
–
106 Hammer2.1-1.5BFunction callingMadeAgents
0.0%
–
106 Hammer2.1-7BFunction callingMadeAgents
0.0%
–
106 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

Memory, 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)Gemma 3 12Bo3Claude Sonnet 4.5Claude 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.

AnthropicGoogleZ.aiDeepSeekxAIOpenAI

20%40%60%80% Jul 20242025Jul2026Jul AnthropicGoogleZ.aiDeepSeekxAIOpenAI

What it measures

Storing and recalling facts across sessions with memory tools.

What it does not measure

Not long-term personal memory in a product: sessions are BFCL's scripted ones.

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