Test and replace AI model#

Click the test on the left. You can test the customer service of the store here until you are satisfied with the test before going online to the front desk of the store.

There are 3 models to choose from

GPT3.5#

Use of fewer tokens, reasonable pricing and fast response speed. GPT-3.5, a transition model between GPT-3 and GPT-4, maintained GPT-3's strengths, but incorporated enhancements in language understanding and generation, offering users more coherent and contextually appropriate responses.

GPT4#

Smarter, uses 10x tokens comparing to GPT3.5. For example, a single conversation in GPT-3.5 consumes about 2k tokens, while the same conversation in GPT-4 would consume approximately 20k tokens.GPT-4, released later, represents a significant leap forward in AI capability. With more training data and improved architecture, GPT-4 can comprehend complex contexts better and produce more nuanced and creative outputs.While it continues to excel in language tasks, GPT-4 also demonstrates impressive capacity for reasoning and problem-solving, pushing the boundary of what AI can achieve.

Claude3 Opus#

Newest model of Claude! The number of tokens used is higher, uses 10x tokens comparing to GPT3.5. Claude3 combines a broad knowledge with strong reasoning and language understanding abilities. This lets it communicate in a clear, coherent way and even tackle complex problems. Claude3's language style is more relaxed and humorous, which is noticeably different from the GPT series.