GPT-4 Turbo on ChatGPT Plus and Team has a 32K token context limit. Is there a specific token limit on input and output?
#GPT-4 input/output token limit on ChatGPT Plus/Team
worldly sapphire
rough coveBOT
Hi @worldly sapphire. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
Based on the information provided in the extracts, GPT-4 Turbo models have a context window of 128,000 tokens. However, the output limit for these models is 4,096 tokens. This means that while the model can receive up to 128K tokens as input, it can only output up to 4,096 tokens.
For instance, the gpt-4-1106-preview model, which is a GPT-4 Turbo model, has a context window of 128,000 tokens and a maximum output of 4,096 tokens. Similarly, the gpt-4-0125-preview model, which is the latest GPT-4 Turbo model, also has a context window of 128,000 tokens and a maximum output of 4,096 tokens.
However, it's important to note that the rate limiter considers both input and output tokens. For example, at tier 1 (paying less than $50 in the past), the gpt-4-turbo-preview has a limit of 150,000 tokens per minute and a more restrictive 500,000 tokens per day.
Please note that these limits are for the API usage and may vary for ChatGPT Plus and Team. For more accurate information, you may want to refer to the specific documentation or support for ChatGPT Plus and Team.
Sources:
- Community Post: Inputs tokens limit, data extraction > Official Answer (https://community.openai.com/t/612242)
- Docs > Models > GPT-4 and GPT-4 Turbo Part 1 (https://platform.openai.com/docs/models)
- Forum Post: Gpt-3.5-turbo-1106 real max tokens > Official Answer (https://community.openai.com/t/589155)
- Forum Post: GPT-4 128K only has 4096 completion tokens > Original post (https://community.openai.com/t/515790)