GPT-4 Turbo
An enhanced model in OpenAI's GPT-4 series, featuring a larger context window and a more recent knowledge cutoff, with performance improvements over the original GPT-4.
Core Facts
The context window in the GPT-3 era was only 2,048 tokens, whereas Google Gemini already supports millions of tokens, Claude supports 200K tokens, and GPT-4 Turbo supports 128K tokens.
GPT-4 Turbo has a context window of 128K tokens.
GPT-4 Turbo has a context window of 128K.
Timeline (last 90 days)
The first DevDay released products including GPT-4 Turbo, the Assistants API, and GPTs custom applications.
The last time OpenAI halted subscription sales was during the GPT-4 Turbo era in 2023.
Llama 3.1 405B's score on MMLU is already close to the level of GPT-4o, and Qwen 2.5 72B has surpassed GPT-4 Turbo on certain metrics in math and code generation benchmarks.
During the transition from GPT-4 to GPT-4 Turbo, OpenAI received a large volume of developer complaints caused by output drift.
2023 年曾有多起 OpenAI API 用户报告 GPT-4 Turbo 的 Token 计数与预期不符，最终确认是系统提示词被重复计入的 bug
All Facts (10)
The context window in the GPT-3 era was only 2,048 tokens, whereas Google Gemini already supports millions of tokens, Claude supports 200K tokens, and GPT-4 Turbo supports 128K tokens.
90%VerifiedGPT-4 Turbo has a context window of 128K tokens.
90%VerifiedGPT-4 Turbo has a context window of 128K.
80%UnverifiedOpenAI于2023年底推出的GPT-4 Turbo对Function Calling的大幅增强降低了Agent类产品的开发门槛
90%UnverifiedLlama 3.1 405B's score on MMLU is already close to the level of GPT-4o, and Qwen 2.5 72B has surpassed GPT-4 Turbo on certain metrics in math and code generation benchmarks.
60%Unverified一个中等复杂度的全栈应用生成可能消耗10万tokens以上，原始API成本约1-3美元
60%UnverifiedThe last time OpenAI halted subscription sales was during the GPT-4 Turbo era in 2023.
50%UnverifiedThe first DevDay released products including GPT-4 Turbo, the Assistants API, and GPTs custom applications.
50%UnverifiedDuring the transition from GPT-4 to GPT-4 Turbo, OpenAI received a large volume of developer complaints caused by output drift.
50%Unverified2023 年曾有多起 OpenAI API 用户报告 GPT-4 Turbo 的 Token 计数与预期不符，最终确认是系统提示词被重复计入的 bug
50%