Unverified50% confidenceFactExact time
Tool-call formats differ across model families: OpenAI uses JSON with a function_call field, Anthropic Claude uses <tool_use> tags, Llama uses <|python_tag|>, and Qwen uses <tool_call> tags.
1
Sources
50%
Confidence
Long-term
Relevance
8/10/2026
First Seen
Sources
DeepSeek V4 Flash工具调用死循环问题解析与解决方案
Related Claims
UnverifiedDifferent models use different function-calling formats: OpenAI uses the tools parameter, Anthropic uses tool_use blocks, and open-source models each have their own template formats.78% similarUnverified不同Provider对工具调用的消息格式存在差异：OpenAI将工具定义放在顶层tools字段并通过role:tool返回结果，Anthropic Claude使用tool_use和tool_result内容块，Google Gemini采用functionDeclarations和functionResponse73% similarUnverifiedOpenAI's Code Interpreter, LangChain's tool-calling framework, and Anthropic's Tool Use API designed for Claude are all representative implementations of the 'LLM + solver' hybrid architecture.71% similarVerifiedOpenAI的Function Calling机制、Anthropic的Tool Use API以及开源框架LangChain、AutoGen均属于工具调用型智能体范畴70% similarUnverifiedOpenAI supports parallel_tool_calls for invoking multiple tools in parallel; Anthropic's Claude uses special <function_calls> XML tags to delimit tool call output; and Google Gemini declares tools through FunctionDeclaration objects.70% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/725463API
curl https://kongchang.com/api/v1/knowledge/claims/725463MCP
get_claim(id=725463)