# Gemini 1.5 pro does not want to write code at full capacity using all its tokens. instead, he shortens the code with comments, which is not the case in gemini 2.0

**URL:** <https://discuss.ai.google.dev/t/gemini-1-5-pro-does-not-want-to-write-code-at-full-capacity-using-all-its-tokens-instead-he-shortens-the-code-with-comments-which-is-not-the-case-in-gemini-2-0/61123>\
**Category:** Google AI Studio\
**Tags:** gemini-15, ai-studio, api, models, gemini-20\
**Created:** [January 16, 2025, 4:42pm UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-does-not-want-to-write-code-at-full-capacity-using-all-its-tokens-instead-he-shortens-the-code-with-comments-which-is-not-the-case-in-gemini-2-0/61123 "2025-01-16T16:42:11Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Earn\_Devison](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/earn_devison/32/38108_2.png) [@Earn\_Devison](https://discuss.ai.google.dev/u/Earn_Devison)\
**Post date:** [January 16, 2025, 4:42pm UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-does-not-want-to-write-code-at-full-capacity-using-all-its-tokens-instead-he-shortens-the-code-with-comments-which-is-not-the-case-in-gemini-2-0/61123/1 "2025-01-16T16:42:12Z")

</div>

Gemini 1.5 pro does not want to write code at full capacity using all its tokens. instead, he shortens the code with comments, which is not the case in gemini 2.0.

---

<div class="post-metadata">

**Author:** ![Dawid\_M](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/dawid_m/32/23772_2.png) [@Dawid\_M](https://discuss.ai.google.dev/u/Dawid_M)\
**Post date:** [January 16, 2025, 6:40pm UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-does-not-want-to-write-code-at-full-capacity-using-all-its-tokens-instead-he-shortens-the-code-with-comments-which-is-not-the-case-in-gemini-2-0/61123/2 "2025-01-16T18:40:03Z")

</div>

I use system instructions like ‘Provide the full code, without gaps, ready to copy.’, and in this way, it gives me complete codes.

---

<div class="post-metadata">

**Author:** ![Earn\_Devison](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/earn_devison/32/38108_2.png) [@Earn\_Devison](https://discuss.ai.google.dev/u/Earn_Devison)\
**Post date:** [January 16, 2025, 9:11pm UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-does-not-want-to-write-code-at-full-capacity-using-all-its-tokens-instead-he-shortens-the-code-with-comments-which-is-not-the-case-in-gemini-2-0/61123/3 "2025-01-16T21:11:49Z")

</div>

It is malfunctioning. Perhaps it functions intermittently; have you experimented with a more extensive codebase?

---

<div class="post-metadata">

**Author:** ![Anthony\_Maio](https://d1dlmcr85iqnpo.cloudfront.net/user_avatar/discuss.ai.google.dev/anthony_maio/32/38561_2.png) [@Anthony\_Maio](https://discuss.ai.google.dev/u/Anthony_Maio)\
**Post date:** [January 21, 2025, 3:27am UTC](https://discuss.ai.google.dev/t/gemini-1-5-pro-does-not-want-to-write-code-at-full-capacity-using-all-its-tokens-instead-he-shortens-the-code-with-comments-which-is-not-the-case-in-gemini-2-0/61123/5 "2025-01-21T03:27:25Z")

</div>

Improve your prompt engineering.