[Feature Request] Dynamic Max Token Limits #883

New Issue

2025-11-11T14:32:59-06:00

GiteaMirror commented

2025-11-11 14:32:59 -06:00

Originally created by @fitchmultz on GitHub (May 10, 2024).

Is your feature request related to a problem? Please describe.
When working with large models, I often need to adjust the token limit based on specific requirements or constraints. The current interface only allows for a fixed maximum token limit of 16k.

Describe the solution you'd like
I'd like a dynamic max token parameter setting that automatically adjusts the maximum token limit based on the selected model. For example, GPT-4-Turbo's context limit is up to 128k, so the max token limit would update accordingly.

Bottom line: Allow users to dynamically set the max token limit based on the selected model for optimal performance and flexibility.

Describe alternatives you've considered
Using my own webui that is forked from an abandoned project the allows setting the context limit to 128000 for GPT-4 models.

Additional context
Having a dynamic max token parameter setting would greatly improve usability and experimentation with large models like GPT-4-Turbo.

Originally created by @fitchmultz on GitHub (May 10, 2024). **Is your feature request related to a problem? Please describe.** When working with large models, I often need to adjust the token limit based on specific requirements or constraints. The current interface only allows for a fixed maximum token limit of 16k. **Describe the solution you'd like** I'd like a dynamic max token parameter setting that automatically adjusts the maximum token limit based on the selected model. For example, GPT-4-Turbo's context limit is up to 128k, so the max token limit would update accordingly. **Bottom line:** Allow users to dynamically set the max token limit based on the selected model for optimal performance and flexibility. **Describe alternatives you've considered** Using my own webui that is forked from an abandoned project the allows setting the context limit to 128000 for GPT-4 models. **Additional context** Having a dynamic max token parameter setting would greatly improve usability and experimentation with large models like GPT-4-Turbo.

GiteaMirror closed this issue

2025-11-11 14:32:59 -06:00

GiteaMirror referenced this issue

2025-11-11 17:22:54 -06:00

[PR #883] [MERGED] fix: error handling #7305

GiteaMirror referenced this issue

2026-04-20 03:00:50 -05:00

[PR #883] [MERGED] fix: error handling #20509

GiteaMirror referenced this issue

2026-04-25 10:08:56 -05:00

[PR #883] [MERGED] fix: error handling #36139

GiteaMirror referenced this issue