🐫 CAMEL: The first and the best multi-agent framework. Finding the Scaling Law of Agents. https://www.camel-ai.org
Python agentai-societiesartificial-intelligencecommunicative-aicooperative-aideep-learninglarge-language-modelsmulti-agent-systemsnatural-language-processing
feat: support gemini 3 (#3433)
Wendong-Fan863887089337
diff --git a/camel/types/enums.py b/camel/types/enums.pyindex ab30d8c4a..79d41ecc7 100644--- a/camel/types/enums.py+++ b/camel/types/enums.py@@ -240,6 +240,7 @@ class ModelType(UnifiedModelType, Enum):NVIDIA_LLAMA3_3_70B_INSTRUCT = "meta/llama-3.3-70b-instruct"# Gemini models+ GEMINI_3_PRO = "gemini-3-pro-preview"GEMINI_2_5_FLASH = "gemini-2.5-flash"GEMINI_2_5_PRO = "gemini-2.5-pro"GEMINI_2_0_FLASH = "gemini-2.0-flash"@@ -835,6 +836,7 @@ class ModelType(UnifiedModelType, Enum):bool: Whether this type of models is gemini."""return self in {+ ModelType.GEMINI_3_PRO,ModelType.GEMINI_2_5_FLASH,ModelType.GEMINI_2_5_PRO,ModelType.GEMINI_2_0_FLASH,@@ -1529,6 +1531,7 @@ class ModelType(UnifiedModelType, Enum):}:return 512_000elif self in {+ ModelType.GEMINI_3_PRO,ModelType.GEMINI_2_5_FLASH,ModelType.GEMINI_2_5_PRO,ModelType.GEMINI_2_0_FLASH,diff --git a/examples/models/gemini_model_example.py b/examples/models/gemini_model_example.pyindex 0cc178358..aff7c44bd 100644--- a/examples/models/gemini_model_example.py+++ b/examples/models/gemini_model_example.py@@ -19,122 +19,49 @@ from camel.types import ModelPlatformType, ModelType# Define system messagesys_msg = "You are a helpful assistant."-# User messageuser_msg = """Say hi to CAMEL AI, one open-source community dedicated to thestudy of autonomous and communicative agents."""--# Example of using the gemini-2.5-pro-preview-06-05 model-model_2_5_pro_pre = ModelFactory.create(+# Example of using the gemini-3-pro-preview model+model_3_pro_pre = ModelFactory.create(model_platform=ModelPlatformType.GEMINI,- model_type=ModelType.GEMINI_2_5_PRO,+ model_type=ModelType.GEMINI_3_PRO,model_config_dict=GeminiConfig(temperature=0.2).as_dict(),)-camel_agent_pro = ChatAgent(system_message=sys_msg, model=model_2_5_pro_pre)+camel_agent_pro = ChatAgent(system_message=sys_msg, model=model_3_pro_pre)response_pro = camel_agent_pro.step(user_msg)print(response_pro.msgs[0].content)--'''-===============================================================================-Hello and a big hi to the entire CAMEL AI community!--It's fantastic to acknowledge your dedication to-the important and fascinating study of autonomous and communicative agents.-Open-source collaboration is the engine of innovation,-and your work is pushing the boundaries of what's possible in AI.--Keep up the brilliant research and community building-===============================================================================-'''---# Example of using the gemini-1.5-pro model-model = ModelFactory.create(- model_platform=ModelPlatformType.GEMINI,- model_type=ModelType.GEMINI_1_5_PRO,- model_config_dict=GeminiConfig(temperature=0.2).as_dict(),-)--# Set agent-camel_agent = ChatAgent(system_message=sys_msg, model=model)--# Get response information-response = camel_agent.step(user_msg)-print(response.msgs[0].content)'''===============================================================================Hi CAMEL AI! 👋-It's great to see a community dedicated to the fascinating field of autonomous-and communicative agents. I'm excited to see what groundbreaking work you're-doing in this area. Keep up the great work! 🤖-===============================================================================-'''--# Example of using the gemini-2.0-flash-exp model-model_2_0_flash = ModelFactory.create(- model_platform=ModelPlatformType.GEMINI,- model_type=ModelType.GEMINI_2_0_FLASH_EXP,- model_config_dict=GeminiConfig(temperature=0.2).as_dict(),-)-camel_agent_flash = ChatAgent(system_message=sys_msg, model=model_2_0_flash)-response_flash = camel_agent_flash.step(user_msg)-print(response_flash.msgs[0].content)--'''-===============================================================================-Hello! I'm happy to say hi to CAMEL AI, one open-source community dedicated to-the study of autonomous and communicative agents. It sounds like a fascinating-community!+It is a pleasure to greet a community so dedicated to the fascinating world of+**autonomous and communicative agents**. Keep up the amazing work pushing the+boundaries of multi-agent collaboration and role-playing frameworks! 🐫🤖==============================================================================='''-# Example of using the gemini-2.0-flash-thinking model-model_2_0_flash_thinking = ModelFactory.create(- model_platform=ModelPlatformType.GEMINI,- model_type=ModelType.GEMINI_2_0_FLASH_THINKING,- model_config_dict=GeminiConfig(temperature=0.2).as_dict(),-)-camel_agent_thinking = ChatAgent(- system_message=sys_msg, model=model_2_0_flash_thinking-)-response_thinking = camel_agent_thinking.step(- "How many rs are there in 'starrary'?"-)-print(response_thinking.msgs[0].content)-'''-===============================================================================-Let's count them out!-s - no r-t - no r-a - no r-r - yes, that's one!-r - yes, that's two!-a - no r-r - yes, that's three!-y - no r--There are **three** rs in "starrary".-===============================================================================-'''---# Example of using the gemini-2.0-pro model-model_2_0_pro = ModelFactory.create(+# Example of using the gemini-2.5-pro model+model_2_5_pro_pre = ModelFactory.create(model_platform=ModelPlatformType.GEMINI,- model_type=ModelType.GEMINI_2_0_PRO_EXP,+ model_type=ModelType.GEMINI_2_5_PRO,model_config_dict=GeminiConfig(temperature=0.2).as_dict(),)-camel_agent_pro = ChatAgent(system_message=sys_msg, model=model_2_0_pro)+camel_agent_pro = ChatAgent(system_message=sys_msg, model=model_2_5_pro_pre)response_pro = camel_agent_pro.step(user_msg)print(response_pro.msgs[0].content)+'''===============================================================================-Hello CAMEL AI! It's great to connect with an open-source community focused on-the exciting field of autonomous and communicative agents. I'm very interested-in learning more about your work and contributions to this area of research.-Best of luck with your endeavors!+Hello and a big hi to the entire CAMEL AI community!++It's fantastic to acknowledge your dedication to+the important and fascinating study of autonomous and communicative agents.+Open-source collaboration is the engine of innovation,+and your work is pushing the boundaries of what's possible in AI.++Keep up the brilliant research and community building==============================================================================='''