Format Sensitivity

BFCL · 50 models

Rank	Model	Score
1	Phi-4 (Prompt)	81.5
2	ToolACE-2-8B (FC)	81.5
3	CoALM-8B	79
4	Llama-3.1-8B-Instruct (Prompt)	74.5
5	GPT-4.1-nano-2025-04-14 (Prompt)	73
6	CoALM-70B	72
7	Gemma-3-4b-it (Prompt)	69.5
8	MiniCPM3-4B (Prompt)	68
9	Gemma-3-12b-it (Prompt)	67.5
10	Claude-Haiku-4-5-20251001 (Prompt)	67.5
11	RZN-T (Prompt)	63.5
12	Qwen3-0.6B (Prompt)	60.5
13	Mistral-Small-2506 (Prompt)	50
14	GPT-4.1-mini-2025-04-14 (Prompt)	45
15	Claude-Sonnet-4-5-20250929 (Prompt)	37.5
16	Bielik-11B-v2.3-Instruct (Prompt)	35
17	Gemma-3-27b-it (Prompt)	34
18	Gemma-3-1b-it (Prompt)	25.5
19	Gemini-2.5-Flash-Lite (Prompt)	25.5
20	GPT-4.1-2025-04-14 (Prompt)	23.5
21	Mistral-Medium-2505	21.5
22	Qwen3-4B-Instruct-2507 (Prompt)	18
23	Qwen3-8B (Prompt)	16.5
24	Qwen3-30B-A3B-Instruct-2507 (Prompt)	16
25	GPT-5-mini-2025-08-07 (Prompt)	16
26	Qwen3-32B (Prompt)	15.5
27	Open-Mistral-Nemo-2407 (Prompt)	14.5
28	Qwen3-14B (Prompt)	14
29	mistral-large-2411 (Prompt)	13.5
30	Claude-Opus-4-5-20251101 (Prompt)	13
31	Grok-4-0709 (Prompt)	13
32	GPT-5.2-2025-12-11 (Prompt)	13
33	DeepSeek-V3.2-Exp (Prompt + Thinking)	10
34	o4-mini-2025-04-16 (Prompt)	9.5
35	Gemini-2.5-Flash (Prompt)	9
36	Gemini-3-Pro-Preview (Prompt)	8.5
37	o3-2025-04-16 (Prompt)	8.5
38	GPT-5-nano-2025-08-07 (Prompt)	8.5
39	Qwen3-235B-A22B-Instruct-2507 (Prompt)	8
40	Ministral-8B-Instruct-2410 (FC)	0
41	GPT-4.1-nano-2025-04-14 (FC)	—
42	palmyra-x-004 (FC)	—
43	Falcon3-3B-Instruct (FC)	—
44	Qwen3-32B (FC)	—
45	xLAM-2-70b-fc-r (FC)	—
46	Granite-3.1-8B-Instruct (FC)	—
47	Mistral-small-2506 (FC)	—
48	Nanbeige3.5-Pro-Thinking (FC)	—
49	DeepSeek-V3.2-Exp (FC)	—
50	Arch-Agent-1.5B	—