Feature Request: Cost Reduction \ Chat Context #850

Closed
opened 2025-11-11 14:32:25 -06:00 by GiteaMirror · 3 comments
Owner

Originally created by @spammenotinoz on GitHub (May 8, 2024).

Brilliant project, but currently using another project as when using this one my API costs are considerably higher.

High API Usage Costs with Some Users who do not appreciate how API's work.

A clear and concise description of what you want to happen.

  • Ability to truncate conversation history by (n) messages. eg: 10. This avoids the entire conversation that may contain irrelevant information back to the LLM. Typically results is a large speed boost and cost reduction due to significant token reduction.

Describe alternatives you've considered

  • User education (training my children) to regularly start new conversations
  • Training users to start a new conversation when finished with embeddings.

Additional context
No modifications to GUI

Originally created by @spammenotinoz on GitHub (May 8, 2024). Brilliant project, but currently using another project as when using this one my API costs are considerably higher. High API Usage Costs with Some Users who do not appreciate how API's work. A clear and concise description of what you want to happen. - Ability to truncate conversation history by (n) messages. eg: 10. This avoids the entire conversation that may contain irrelevant information back to the LLM. Typically results is a large speed boost and cost reduction due to significant token reduction. **Describe alternatives you've considered** - User education (training my children) to regularly start new conversations - Training users to start a new conversation when finished with embeddings. **Additional context** No modifications to GUI
Author
Owner

@justinh-rahb commented on GitHub (May 8, 2024):

I too had some expensive months due to people putting everything in the same conversation. The solution was user training: telling them to start new threads for new topics.

@justinh-rahb commented on GitHub (May 8, 2024): I too had some expensive months due to people putting everything in the same conversation. The solution was user training: telling them to start new threads for new topics.
Author
Owner

@spammenotinoz commented on GitHub (May 12, 2024):

My bad, this feature is already in "Advanced".

@spammenotinoz commented on GitHub (May 12, 2024): My bad, this feature is already in "Advanced".
Author
Owner

@GrayXu commented on GitHub (Aug 1, 2024):

@spammenotinoz
hi, how do you make modifications in "Advanced"? It seems there's only Context Length related to this, but even when I set it to 1, I can see the full context being used for each interaction (the single token consumption keeps increasing).
thanks!

@GrayXu commented on GitHub (Aug 1, 2024): @spammenotinoz hi, how do you make modifications in "Advanced"? It seems there's only `Context Length` related to this, but even when I set it to 1, I can see the full context being used for each interaction (the single token consumption keeps increasing). thanks!
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: github-starred/open-webui#850