GPT-4
GPT-4 is a text-based AI model by OpenAI. It features a 8K token context window, 33 tok/s output speed, $30/1M input tokens pricing. It scores 12.8 on the intelligence index.
Weaknesses
- -Slower than average output (33 tok/s)
- -Limited context window (8K)
- -Premium pricing
Benchmarks
Intelligence12.8Coding13.1MathN/A
Intelligence
12.8
Coding
13.1
Output Speed
33 tok/s
Latency (TTFT)
0.74s
Blended Price
$37.50/M
Pricing per 1M Tokens
| Input (Prompt) | $30 |
| Output (Completion) | $60 |
| Cache Read | Free |
| Cache Write | Free |
Specifications
| Context Length | 8K |
| Max Output Tokens | N/A |
| Input Modalities | Text |
| Output Modalities | Text |
| Tokenizer | N/A |
Compare GPT-4 with:
Alternatives to GPT-4
More from OpenAI
Frequently Asked Questions
How much does GPT-4 cost?
GPT-4 costs $30/1M input tokens and $60/1M output tokens.
What is GPT-4's context window?
GPT-4 supports a context window of 8K tokens, which is approximately 4 pages of text.
Is GPT-4 good for coding?
GPT-4 has a lower coding score of 13.1. For demanding coding tasks, consider a model with a higher coding benchmark.
How fast is GPT-4?
GPT-4 generates output at 33 tok/s. Time to first token is 0.74s.
Is there a free version of GPT-4?
No, GPT-4 is a paid model. Check the free models page for zero-cost alternatives.
Last updated: