OpenAI: GPT-4

Text input Text output
Author's Description

OpenAI's flagship model, GPT-4 is a large-scale multimodal language model capable of solving difficult problems with greater accuracy than previous models due to its broader general knowledge and advanced reasoning capabilities. Training data: up to Sep 2021.

Key Specifications
Cost
+$$$$$
Context
8K
Released
May 27, 2023
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Tool Choice Top P Logit Bias Temperature Logprobs Presence Penalty Stop Response Format Structured Outputs Tools Max Tokens Frequency Penalty Top Logprobs Seed
Features

This model supports the following features:

Response Format Tools Structured Outputs
Performance Summary

OpenAI's GPT-4, released on May 27, 2023, demonstrates competitive response times, ranking in the 41st percentile across evaluated benchmarks. However, it is positioned at a premium price point, falling into the 4th percentile for cost-effectiveness. The model's reliability information was not provided, so no specific comment can be made on this aspect. GPT-4 exhibits exceptional performance in Ethics, achieving perfect 100.0% accuracy, making it the most accurate model at its price point and among models with comparable speed. It also shows strong capabilities in Instruction Following, with a 67.0% accuracy, placing it in the 75th percentile for this category. Conversely, its performance in Coding is a notable weakness, with only 7.0% accuracy, ranking in the 13th percentile. This suggests that while GPT-4 excels in ethical reasoning and understanding complex instructions, its programming knowledge requires further development. Its strengths lie in its advanced reasoning and general knowledge, as evidenced by its high accuracy in ethics and instruction following.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $30
Completion $60

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
OpenAI
OpenAI | openai/gpt-4 8K $30 / 1M tokens $60 / 1M tokens
Azure
Azure | openai/gpt-4 8K $30 / 1M tokens $60 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Strategy Free Executions Accuracy Cost Duration
Other Models by openai