What is the character limit for ChatGPT? Explanation of how to avoid it and upper limits by model

When using ChatGPT, have you ever had experiences such as, "The output stopped halfway through" or "Even though I specified the number of characters, it didn't output properly"?
This model is very groundbreaking and not only allows for natural conversations, but also allows for email creation, summary papers, and even generate creative sentences such as novels. However, one of the weaknesses of ChatGPT is that if the text is too long, input and output will not work properly .

Click here for more information about OpenAI's latest AI agent, OpenAI Deep Research ⬇️
[ChatGPT] What is OpenAI Deep Research? A thorough explanation of how to use it and the fee structure!

table of contents

What is the character limit in ChatGPT?

Character limit for input to ChatGPT

There is a limit to answers from ChatGPT

How to remove the character limit in ChatGPT

When input: Enter multiple times

Output 1: Send instructions to divide and output

Output time 2: If generation stops halfway through

Common: Translate into English and input/output

Common: Use ChatGPT via API

Why ChatGPT has a character limit

Language model processing power

Ensuring quality of response

summary

What is the character limit in ChatGPT?

The ChatGPT character limit refers to the maximum number of characters that can be entered in a single request (conversation).
Generally, many articles introduce the maximum number of characters that can be entered into ChatGPT in a single question as approximately 4,096 tokens.

However, OpenAI has not released any clear numbers regarding ChatGPT character limits. Therefore, there are some differences in opinion regarding the exact maximum number of characters.
Furthermore, there is no clear standard for the output limit, and there are many cases where "I want to output long sentences, but the generation stops halfway through."

So, this time we have examined how many characters you can actually enter, both the free and paid versions, so we will explain based on the results!

!

A token is the unit used by ChatGPT when processing sentences. In Japanese, it is assumed that one tokens per character is 1 to 2 tokens per character, and in English, it is assumed that one token is one word.

Character limit for input to ChatGPT

Here we have verified the maximum number of characters that can be entered for GPT-3.5 (free version) and GPT-4 (paid version: ChatGPT Plus), and introduce the tables.

As a prerequisite for verification, the following conditions were carried out:

• Using Bitcoin papers
• Before the text, place the prompt "Please summarize this sentence."
• Use the same sentences in both Japanese and English

The results are as follows:

Model	Japanese input limit	English input limit
GPT3.5	Approximately 6,000 characters	Approximately 40,000 characters
GPT4	Approximately 22,000 characters	Approximately 50,000 characters

Thus, the maximum number of characters (tokens) was very different depending on GPT-3.5 and GPT-4, and whether it's Japanese or English.
In Japanese, there is a significant difference in the maximum character limit, but in English, there is no significant difference between the free version and the paid version .

!

The above results are guidelines, and the method of calculating the number of tokens will vary depending on the language and the content of the sentence (such as the number of kanji and context). Please consider this as a reference only.

In addition, the character limit for GPT-4o is explained in the article below.
➡️What is ChatGPT-4o (GPT4 omni)? A thorough explanation of how to use it, prices, and more!

There is a limit to answers from ChatGPT

The character limit in ChatGPT is not only limited to input limits, but also for answering questions.

Since it is difficult to verify the output limit, I asked GPT3.5 and GPT4 instead, "What is the maximum number of characters you output?" and the results were as follows:

Model	Output limit
GPT-3.5	Approximately 2048 characters
GPT-4	Approximately 3000-4000 characters

!

The output limit is only a guideline, and there may be variations in output accuracy and volume of sentences depending on the timing of use (if the time of use is or if the server is busy).

In addition, it is not only the maximum number of characters to be output, but it also appears that the output does not output as specified .

Even in GPT4, ChatGPT seems to be weak to character counting, and it is difficult to meet the lower limit of character limit, but if anyone has come up with a prompt to solve it, please let me know.

— KAJI | Kajitani Kento (@kajikent) April 5, 2023

In this way, even if you specify the lower limit of the number of characters, "more than 100 characters", there are many cases where it is not possible to output the number of characters as specified.

In fact, I tried to output the report using GPT3.5 and GPT4 with the instructions that "Please output at least 1,500 characters" on the lower birthrate and aging population.

No matter how many times I output, only about 1,000 characters were output, and the result was that output was not possible that satisfies the minimum number of characters specified.

How to remove the character limit in ChatGPT

As mentioned above, character limits are inevitable when using ChatGPT.
However, by devising the prompt , it is possible to avoid the maximum number of characters and input/output .

Here are some ways to avoid the limitations.

When input: Enter multiple times

For long sentences, you can simply enter them in a split format rather than just filling them all at once.

Below is an example prompt:

<code>今から送る文章を要約して下さい。以下がその手順です。 ###ステップ1:文章の量が長いので3回に分けて送ります。ステップ2:全て送り終わったら、"完了"と伝えます。ステップ3私が"完了"と伝えるまで、あなたは"了解しました"とだけ答えて下さい。 "完了"と言われたら、要約を作成してください。</code>

By issuing such instructions, even sentences that exceed the maximum character limit can be loaded into ChatGPT.

!

Input is possible, but if the total amount of text is too large, the accuracy of the answer may be reduced due to reasons such as accuracy and limits to understanding the context. please note.

Output 1: Send instructions to divide and output

When outputting long sentences, such as writing reports or blog articles, it often doesn't work even if you try to output them all at once.
Therefore, by giving instructions to output the text divided into sections and characters , you can output long sentences more accurately and as desired.

Below is an example prompt:

<code>今からブログ記事の草案を送ります。その草案に沿って、"記事全体で2000文字以上”になるように本文を作成して下さい。条件として、一度に出力するのではなく、見出しごとに作成を行なって下さい。そして、次の見出しに進む前に私に確認をとって下さい。記事構成は以下の通りです。 ###タイトル:ダイエット成功の秘訣見出し:はじめにダイエットの目的と重要性の概説。健康的な食事計画の要点。効果的な運動の種類とその組み合わせ。ダイエット時のモチベーション維持の方法。ストレス管理の重要性。まとめ###</code>

I actually created a blog post using the prompts above.

Examples created for each section

In this way, by adding an instruction to the prompt to split the output instead of outputting to ChatGPT all at once, you can expect more detailed and careful output.

Output time 2: If generation stops halfway through

When outputting a text, the response may stop halfway through, but in that case, pressing the "Continue Generating" button on the image will continue to output.

Examples of interrupted generation

Common: Translate into English and input/output

As mentioned at the beginning, there is a significant difference in the amount of sentences that can be entered between Japanese and English . This is because even with the same sentence, the amount of tokens in English is less than half.
Therefore, unless there is a special reason such as "I really don't want to distort the meaning of the sentence in Japanese", it is thought that by translating it into English, you can enter more amounts of sentences than in Japanese.

This is the same for output. Especially when using via an API , it is recommended to input and output in English as it reduces costs.

Common: Use ChatGPT via API

Why ChatGPT has a character limit

Many people have probably experienced the inconvenience of using ChatGPT because of character limit errors or the inability to output long sentences.

However, this is necessary for both users and ChatGPT to ensure comfortable use .

There are two main reasons why ChatGPT has a character limit:

Language model processing power

• ChatGPT is a high-performance language model, but there are limits to processing large amounts of text simultaneously, and there are concerns that it will put a strain on the server.

Ensuring quality of response

• It is difficult to generate high-quality, contextual responses for too long text, and there is a risk that the data will be generated in a logically broken state.
• Furthermore, processing text and generating answers can take time, which may lead to user convenience.

In this way, the character limit set in ChatGPT is considered necessary to maintain the processing power of the language model, maintain high-quality responses, and improve the user experience .
It may seem inconvenient, but a certain level of character limit is essential to achieving proper and highly accurate generation.

summary

While we, users are dissatisfied with the limit, ChatGPT's character limit is an important feature in maintaining a balance between user experience and the performance of the language model.
By understanding the difference between character limits between the free and paid versions and paying attention to creating effective prompts, you can make use of ChatGPT more comfortable!

We hope this article will help you to gain a deeper understanding of ChatGPT and effectively incorporate it into your daily activities!

The above is the detailed content of What is the character limit for ChatGPT? Explanation of how to avoid it and upper limits by model. For more information, please follow other related articles on the PHP Chinese website!