GPT-3.5 Turbo (OpenAI)

GPT-3.5 Turbo is OpenAI's conversational large language model released in March 2023 and quickly incorporated into the popular ChatGPT offering. Its defining technical shift from earlier GPT-3 text models is the move away from unstructured token-sequence input toward a message-based interface using Chat Markup Language, commonly called ChatML. This ChatML framing lets the model interpret a series of role-tagged messages rather than a flat string of tokens, which made it a natural fit for chatbot-style applications, instruction-following, and developer APIs that need to model system, user, and assistant turns in a structured way.

Practically, GPT-3.5 Turbo was positioned as a cost-efficient alternative to prior GPT-3 generation endpoints, with OpenAI marketing it as roughly an order of magnitude more affordable than its GPT-3 predecessors. That price-to-capability ratio, combined with the ChatML message format, made the model attractive for teams prototyping chat assistants, summarization tools, retrieval-augmented workflows, and other natural-language features where strong conversational fluency mattered more than cutting-edge reasoning. Its real-world fit is best described as a general-purpose workhorse for chat and short-form generation in production systems, rather than a frontier reasoning model.