TLDR
- OpenAI launched GPT-4o mini, a more affordable and efficient version of GPT-4o
- GPT-4o mini offers comparable performance to larger models at a fraction of the cost
- The new model supports text and vision, with audio and video support coming soon
- GPT-4o mini is available on Azure AI with enhanced safety features
- The model has a 128,000 token context window and is significantly cheaper than previous versions
OpenAI, the US-based artificial intelligence company, has launched a new generative AI model called GPT-4o mini.
This new model aims to provide comparable performance to larger models at a lower cost, potentially expanding the range of AI applications.
GPT-4o mini is essentially a more cost-effective version of OpenAI’s current top-of-the-line consumer model. According to OpenAI, the new model is “an order of magnitude more affordable than previous frontier models” and “more than 60% cheaper than GPT-3.5 Turbo.”
Despite its smaller size and lower energy consumption, GPT-4o mini doesn’t seem to lack in performance. OpenAI states that it surpasses GPT-3.5 Turbo and other small models on academic benchmarks across both textual intelligence and multimodal reasoning. It also supports the same range of languages as GPT-4o.
towards intelligence too cheap to meter:https://t.co/76GEqATfws
15 cents per million input tokens, 60 cents per million output tokens, MMLU of 82%, and fast.
most importantly, we think people will really, really like using the new model.
— Sam Altman (@sama) July 18, 2024
Currently, GPT-4o mini supports text and vision capabilities. OpenAI has announced that support for audio and video inputs and outputs is “coming in the future.”
The model has a context window of 128,000 tokens, which is a measurement of how much it can remember in a given conversation. This is significantly larger than GPT-3.5 Turbo’s 16,000 token context window.
In terms of pricing, GPT-4o mini costs 15 cents per million input tokens and 60 cents per million output tokens. This is substantially cheaper than GPT-4o, which costs $5 per million input tokens and $2.50 per million output tokens.
OpenAI envisions a future where AI models become seamlessly integrated into every app and website. They believe GPT-4o mini is paving the way for developers to build and scale powerful AI applications more efficiently and affordably.
Microsoft’s Azure AI platform has also announced the availability of GPT-4o mini. Azure AI is extending its safety features to the new model, including prompt shields and protected material detection, which are now ‘on by default’ for users of GPT-4o mini on Azure OpenAI Service.
Azure AI is offering GPT-4o mini with data residency options in 27 regions, giving customers control over where their data is stored and processed. This feature aims to help customers meet their unique compliance requirements.
The new model is available on Azure AI’s global pay-as-you-go deployment, allowing customers to pay only for the resources they consume. This option offers higher throughput while still providing control over where data resides at rest.
Azure AI is also introducing GPT-4o mini to its Batch service, which delivers high-throughput jobs with a 24-hour turnaround at a 50% discount rate by using off-peak capacity. Fine-tuning for GPT-4o mini will be available, allowing customers to customize the model for specific use cases.