OpenAI, led by Sam Altman, has introduced a new cost-efficient small AI model named GPT-40 Mini. It is 60% cheaper than GPT-3.5 Turbo, priced at 15 cents per million input tokens and 60 cents per million output tokens.
GPT-40 Mini Features
According to a press release by the Microsoft-backed company, GPT-40 Mini can outperform GPT-41 on chat preferences in the LMSYS leaderboard, scoring 82% on Massive Multitask Language Understanding (MMLU). This model supports a variety of tasks, including applications that chain or parallelize multiple model calls (e.g., calling multiple APIs), pass a large volume of context to the model (e.g., full code base or conversation history), or interact with customers through fast, real-time text responses (e.g., customer support chatbots).
Currently, GPT-40 Mini supports text and vision in the API, with plans to add support for text, image, video, and audio inputs and outputs in the future.
Performance-wise, GPT-40 Mini scored 87.0% in math reasoning, compared to 75.5% for Gemini Flash and 71.7% for Claude Haiku. It also scored 87.2% on HumanEval, which measures coding performance, compared to 71.5% for Gemini Flash and 75.9% for Claude Haiku.
Availability and Pricing
GPT-40 Mini is available as a text and vision model in the Assistants API, Chat Completions API, and Batch API. Developers are charged 15 cents per 1 million input tokens and 60 cents per 1 million output tokens (roughly equivalent to 2500 pages in a standard book).
In ChatGPT, Free, Plus, and Team users can access GPT-40 Mini starting today, replacing GPT-3.5. Enterprise users will have access starting next week.
Safety Measures in GPT-40 Mini
OpenAI ensures that GPT-40 Mini has the same safety mitigations as GPT-40. More than 70 external experts in fields such as social psychology and misinformation have tested GPT-40 to identify potential risks. Insights from these expert evaluations have helped improve the safety of both GPT-40 and GPT-40 Mini.
