Huge ChatGPT Update | Plugin Competitor & 16k GPT-3.5 Turbo
Nick Franck Nick Franck
192 subscribers
1,637 views
0

 Published On Jun 14, 2023

OpenAI released a huge update. You can see the update here: https://openai.com/blog/function-call...

The updates being introduced include:
1. New function calling capability in the Chat Completions API: GPT-4-0613 and GPT-3.5-turbo-0613 can now help decide which functions developers should execute based on user input, allowing GPT's capabilities to be better integrated with external tools and APIs. The models can now output a JSON object containing arguments to call these functions based on user input. This allows for more efficient conversion of natural language into API calls or database queries, and extraction of structured data from text.

2. Updated and more steerable versions of gpt-4 and gpt-3.5-turbo: The newly released models have improved steerability, which enhances developer control over model responses.

3. New 16k context version of gpt-3.5-turbo: This version, priced at twice the standard 4k version, can handle about 20 pages of text in a single request, offering developers greater flexibility in handling larger texts.

4. Cost Reduction: A 75% cost reduction on the state-of-the-art embeddings model and a 25% cost reduction on input tokens for gpt-3.5-turbo have been implemented.

5. Deprecation of older models: There's an announcement of the deprecation timeline for the gpt-3.5-turbo-0301 and gpt-4-0314 models, with new models replacing them. Applications using the stable model names will be automatically upgraded.

All these improvements maintain the data privacy and security standards where customers own all outputs generated from their requests and their API data will not be used for training.

Also, there is an intent to invite more people from the waitlist to try GPT-4 over the coming weeks, with plans to remove the waitlist entirely. The changes reflect continuous efforts to enhance model capabilities and efficiency based on feedback from developers.

00:00 Intro
00:08 Steering
00:53 16k GPT-3.5 Turbo
02:14 16k Pricing
02:58 Embeddings Pricing
04:06 Function Calling
06:26 Model Upgrade/Deprecation

show more

Share/Embed