Tool Calling Model Ranking

AI BENCHY Category

See which AI models perform best on Tool Calling, which ones stay reliable, and where the biggest gaps appear. Sort by: Response Time (avg) ↓.

Models Shown

Average Tool Calling Score

8.7

Best Model

Failure Reasons

With failure reason API error15 With failure reason Invalid tool call7 With failure reason Did not follow instructions6 With failure reason No answer2 With failure reason Wrong answer2

Rank	Model	Company	Tool Calling Score	Score	Tests Correct	Response Time (avg)
#58	Gemini 3.1 Flash Lite Preview none	Google	10.0	7.2	1/1	3.39s
#48	Gemini 3 Flash Preview none	Google	10.0	7.4	1/1	3.35s
#95	Qwen3.5 Plus 2026-02-15 none	Qwen	10.0	6.3	1/1	3.33s
#123	MiMo-V2.5-Pro none	Xiaomi	10.0	5.5	1/1	3.30s
#3	Gemini 3.5 Flash low	Google	10.0	9.4	1/1	3.27s
#57	Step 3.7 Flash low	Stepfun	10.0	7.3	1/1	3.25s
#104	Nemotron 3 Ultra 550b A55b none	NVIDIA	10.0	6.0	1/1	2.99s
#90	Gemini 3.1 Flash Lite none	Google	10.0	6.4	1/1	2.97s
#136	Elephant Alpha medium	Openrouter	3.0	5.1	0/1	2.83s
#71	Step 3.7 Flash high	Stepfun	10.0	7.0	1/1	2.79s
#32	Gemini 3.5 Flash minimal	Google	10.0	7.7	1/1	2.79s
#137	Elephant Alpha none	Openrouter	3.0	5.1	0/1	2.79s
#125	GPT-5.4 none	OpenAI	10.0	5.5	1/1	2.75s
#150	Qwen3 Coder Next medium	Qwen	10.0	4.6	1/1	2.64s
#147	GPT-4o-mini none	OpenAI	10.0	4.8	1/1	2.51s

Tool Calling Ranking