List your company

Agentic Tool Calling on BFCL v3 (Multi-Turn)

60.5Multi-Turn Success RateKimi-K2
32.4239.714754.29May 29, 2026

Evaluation results

MethodLinks
Kimi-K2Model Identifier=Kimi-K22026.0560.5
MAVENBase Model=GPT-OSS-120b2026.0558.5
Qwen3-Th-235BModel Identifier=Qwen3-235B2026.0553.5
o4-miniModel Identifier=o4-mini-2025-04-162026.0553
o3Model Identifier=o3-2025-04-162026.0544
DeepSeek-V3.1Model Identifier=DeepSeek-V3.12026.0544
Gemini-2.5Model Identifier=gemini-2.5-pro-preview-05-062026.0535
GPT-5Model Identifier=gpt-5-2025-08-072026.0533.5