# Gpt-4-vision-preview \`finish\_details\`
**URL:** https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911
**Category:** API
**Created:** [November 6, 2023, 10:09pm UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911 "2023-11-06T22:09:47Z")
**Posts on this page:** 10
**Page:** 1
### Author:  [@dr.ahfz](https://community.openai.com/u/dr.ahfz)
#### Post date: [November 6, 2023, 10:09pm UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911/1 "2023-11-06T22:09:47Z")
I’m trying out the `gpt-4-vision-preview` and no matter what I do I get back truncated output at around 50 char. The `finish_details` field says `max_tokens`. However in the response details I can see I am well below the maximum tokens for this request. Anyone else have this issue?
---
### Author:  [@lucasrantzau](https://community.openai.com/u/lucasrantzau)
#### Post date: [November 6, 2023, 10:15pm UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911/2 "2023-11-06T22:15:32Z")
Yes, I am experiencing the same issue. Might it be a limitation now while they roll it out?
---
### Author:  [@dr.ahfz](https://community.openai.com/u/dr.ahfz)
#### Post date: [November 6, 2023, 10:46pm UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911/3 "2023-11-06T22:46:44Z")
Setting `max_tokens` in the request gets around the issue.
---
### Author:  [@lucasrantzau](https://community.openai.com/u/lucasrantzau)
#### Post date: [November 6, 2023, 11:00pm UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911/4 "2023-11-06T23:00:30Z")
I have set “max\_tokens: 4096,” which is the maximum, but it’s still cutting the response off at around 50 characters. How strange. What have you set it to?
---
### Author:  [@dr.ahfz](https://community.openai.com/u/dr.ahfz)
#### Post date: [November 6, 2023, 11:17pm UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911/5 "2023-11-06T23:17:19Z")
I set it to 1500, which is enough for me:
```auto
model: 'gpt-4-vision-preview',
max_tokens: 1500,
```
---
### Author:  [@Decrypt](https://community.openai.com/u/Decrypt)
#### Post date: [November 7, 2023, 12:26am UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911/6 "2023-11-07T00:26:16Z")
omg thank you, I was going crazy.
Using :
```auto
if (currentModel === 'gpt-4-vision-preview') {
requestPayload.max_tokens = 4096;
}
```
fixed my discord bot. I don’t know why, by default it seems to be at 16 tokens, which is kinda ridiculous.
---
### Author:  [@baki.kucukcakiroglu](https://community.openai.com/u/baki.kucukcakiroglu)
#### Post date: [November 7, 2023, 6:06am UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911/7 "2023-11-07T06:06:47Z")
I also experienced an issue similar to this. Reponses of `gpt-4-vision-preview`was truncated to 31 tokens with no reason and finish details type was ‘max\_tokens’. I set ‘max\_tokens’ parameter of request to 4096 and it works well now.
---
### Author:  [@ejeyyvbpp](https://community.openai.com/u/ejeyyvbpp)
#### Post date: [November 17, 2023, 7:32am UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911/9 "2023-11-17T07:32:33Z")
i think this is a bug, There is a problem with the default values of this model
---
### Author:  [@ultrasorak](https://community.openai.com/u/ultrasorak)
#### Post date: [November 21, 2023, 4:40am UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911/10 "2023-11-21T04:40:00Z")
Any update on this issue? I am also facing the same.
---
### Author:  [@\_j](https://community.openai.com/u/_j)
#### Post date: [November 21, 2023, 4:49am UTC](https://community.openai.com/t/gpt-4-vision-preview-finish-details/475911/11 "2023-11-21T04:49:26Z")
The update is documentation:
[https://platform.openai.com/docs/guides/vision](https://platform.openai.com/docs/guides/vision)
> Currently, GPT-4 Turbo with vision does not support the **message.name** parameter, **functions/tools** , **response\_format** parameter, and we currently set a low **max\_tokens** default which you can override.
Why do they set a low default value that could truncate the typical response? Life’s little mysteries.