🚀 OpenAI Launches GPT-4o Mini!

Hey, we know OpenAI has just launched GPT-4o Mini, a model that's 60% cheaper than GPT-3.5 and the best among all the small LLMs. This marks a literal 99% cost reduction in GPT-3 equivalent models over the last two years.

OpenAI has formally acknowledged this cost reduction trajectory:

"The cost per token of GPT-4o Mini has dropped by 99% since text-davinci-003, a less capable model introduced in 2022. We’re committed to continuing this trajectory of driving down costs while enhancing model capabilities."

💰 Is GPT-4o Mini Free?

No, but it's very affordable! OpenAI has made GPT-4o Mini available as a text and vision model immediately in the Assistant API, Chat Completion API, and the Batch API. You only need to pay 15 cents per 1M input prompt tokens and 60 cents per 1M output response tokens.

OpenAI expects GPT-4o Mini to significantly expand the range of applications built with AI by making intelligence much more affordable.

🧙‍♂️ Does Merlin Offer GPT-4o Mini?

Yes! Merlin GPT-4o Mini is now live on Merlin at getmerlin.in/magic. You can access Merlin's unified model and feature router, which detects the appropriate model and Merlin feature to address your query based on the prompt you put in. It can be accessed via the "Merlin Magic" button.

📏 How Big is GPT-4o Mini?

• Context Window: 128K tokens
• Output Tokens: Supports up to 16K output tokens per request
• Knowledge Cutoff: October 2023

GPT-4o Mini's low cost and latency enable a broad range of tasks, such as applications that chain or parallelize multiple model calls (e.g., calling multiple APIs), passing a large volume of context to the model (e.g., full code base or conversation history), or interacting with customers through fast, real-time text responses (e.g., customer support chatbots).

Today, GPT-4o Mini supports text and vision in the API, with future support planned for text, image, video, and audio inputs and outputs. Thanks to the improved tokenizer shared with GPT-4o, handling non-English text is now even more cost-effective.

📊 Performance and Use Cases

A Small Model with Superior Textual Intelligence and Multimodal Reasoning

• Superior Performance:

  • GPT-4o Mini surpasses GPT-3.5 Turbo and other small models on academic benchmarks.
  • Excels in both textual intelligence and multimodal reasoning.

• Language Support:

  • Supports the same range of languages as GPT-4o.

• Function Calling:

  • Demonstrates strong performance in function calling.
  • Enables developers to build applications that fetch data or take actions with external systems.

• Long-Context Performance:

  • Shows improved long-context performance compared to GPT-3.5 Turbo.

🔍 Key Benchmark Evaluations:

• Reasoning Tasks:

  • MMLU: GPT-4o Mini scored 82.0%, compared to 77.9% for Gemini Flash and 73.8% for Claude Haiku.

• Math and Coding Proficiency:

  • MGSM (Math Reasoning): GPT-4o Mini scored 87.0%, compared to 75.5% for Gemini Flash and 71.7% for Claude Haiku.
  • HumanEval (Coding Performance): GPT-4o Mini scored 87.2%, compared to 71.5% for Gemini Flash and 75.9% for Claude Haiku.

• Multimodal Reasoning:

  • MMMU: GPT-4o Mini scored 59.4%, compared to 56.1% for Gemini Flash and 50.2% for Claude Haiku.

🆓 Access in ChatGPT

In ChatGPT Free, Plus, and Team plans, you will be able to access GPT-4o Mini starting today, replacing GPT-3.5.

GPT-4o Mini is now live on Merlin at getmerlin.in/magic.

Feel free to reach out with any questions or for more details on how GPT-4o Mini can benefit your projects!