more than one copilot credit is used for tasks #7562

Closed
opened 2026-02-16 18:07:35 -05:00 by yindo · 10 comments
Owner

Originally created by @inteist on GitHub (Jan 25, 2026).

Originally assigned to: @rekram1-node on GitHub.

Description

when a task is executed via copilot-cli OR copilot VSCode extension, it uses 1 premium credit
when same task is executed via opencode, it can use an arbitrary number of credits - usually 2-3 but can be more depending on the complexity of the task.

since the integration with Copilot is official, opencode should not consume more than 1 premium credit for a single task

Plugins

No response

OpenCode version

No response

Steps to reproduce

No response

Screenshot and/or share link

No response

Operating System

No response

Terminal

No response

Originally created by @inteist on GitHub (Jan 25, 2026). Originally assigned to: @rekram1-node on GitHub. ### Description when a task is executed via copilot-cli OR copilot VSCode extension, it uses 1 premium credit when same task is executed via opencode, it can use an arbitrary number of credits - usually 2-3 but can be more depending on the complexity of the task. since the integration with Copilot is official, opencode should not consume more than 1 premium credit for a single task ### Plugins _No response_ ### OpenCode version _No response_ ### Steps to reproduce _No response_ ### Screenshot and/or share link _No response_ ### Operating System _No response_ ### Terminal _No response_
yindo closed this issue 2026-02-16 18:07:35 -05:00
Author
Owner

@github-actions[bot] commented on GitHub (Jan 25, 2026):

This issue might be a duplicate of existing issues. Please check:

  • #8700: GH Copilot | synthetic user messages burn premium requests
  • #8030: Copilot auth now sets far too many requests as 'user' consuming premium requests rapidly
  • #8067: Github Copilot multiple premium request charges without subagents

All of these issues are related to the same core problem: synthetic user messages and incorrect message role classification causing excessive premium credit consumption. Feel free to ignore if this issue describes a different specific scenario.

@github-actions[bot] commented on GitHub (Jan 25, 2026): This issue might be a duplicate of existing issues. Please check: - #8700: GH Copilot | synthetic user messages burn premium requests - #8030: Copilot auth now sets far too many requests as 'user' consuming premium requests rapidly - #8067: Github Copilot multiple premium request charges without subagents All of these issues are related to the same core problem: synthetic user messages and incorrect message role classification causing excessive premium credit consumption. Feel free to ignore if this issue describes a different specific scenario.
Author
Owner

@rekram1-node commented on GitHub (Jan 25, 2026):

This should not happen, need to know ur opencode version, if you are using plugins, and u would need to share your config too

@rekram1-node commented on GitHub (Jan 25, 2026): This should not happen, need to know ur opencode version, if you are using plugins, and u would need to share your config too
Author
Owner

@OliverDev2k3 commented on GitHub (Jan 25, 2026):

I have same problem
my opencode version 1.1.34

@OliverDev2k3 commented on GitHub (Jan 25, 2026): I have same problem my opencode version 1.1.34
Author
Owner

@rekram1-node commented on GitHub (Jan 25, 2026):

please share the necessary things to debug as requested :)

@rekram1-node commented on GitHub (Jan 25, 2026): please share the necessary things to debug as requested :)
Author
Owner

@ks-extlm commented on GitHub (Jan 26, 2026):

I have the same issues since official copilot plugin was integrated. Sometimes I am being charged 1 premium request per 1 prompt even for models like Sonnet 4.5.

opencode version 1.1.34
plugins: superpowers

@ks-extlm commented on GitHub (Jan 26, 2026): I have the same issues since official copilot plugin was integrated. Sometimes I am being charged 1 premium request per 1 prompt even for models like Sonnet 4.5. opencode version 1.1.34 plugins: superpowers
Author
Owner

@tonthanhhung commented on GitHub (Jan 28, 2026):

I'm facing the same issue, using the opus 4.5 model previously cost 3 requests (which is correct multiplication from their doc) but now with a single task, it ran out all my remaining requests for a month. Which makes it unusable anymore. I can't replicate the issue if you asked for a debug log until next month's budget allocated 😥

@tonthanhhung commented on GitHub (Jan 28, 2026): I'm facing the same issue, using the opus 4.5 model previously cost 3 requests (which is correct multiplication from their doc) but now with a single task, it ran out all my remaining requests for a month. Which makes it unusable anymore. I can't replicate the issue if you asked for a debug log until next month's budget allocated 😥
Author
Owner

@ks-extlm commented on GitHub (Jan 28, 2026):

It might have improved a bit with 1.1.39.

One thing worth mentioning here is that OpenCode uses "cheaper" model to generate session title message. For GitHub Copilot, cheaper means Haiku 4.5 that has 0.33 premium request multiplier. I use GPT-5 mini that has 0 multiplier (it is effectively free for any paid GitHub plan) and it works fine for such simple tasks. Just add this to your configuration:

"small_model": "github-copilot/gpt-5-mini"

@ks-extlm commented on GitHub (Jan 28, 2026): It might have improved a bit with 1.1.39. One thing worth mentioning here is that OpenCode uses "cheaper" model to generate session title message. For GitHub Copilot, cheaper means Haiku 4.5 that has 0.33 premium request multiplier. I use GPT-5 mini that has 0 multiplier (it is effectively free for any paid GitHub plan) and it works fine for such simple tasks. Just add this to your configuration: `"small_model": "github-copilot/gpt-5-mini"`
Author
Owner

@inteist commented on GitHub (Jan 28, 2026):

Doesn't seem to happen anymore on 1.1.40
Thanks!

@inteist commented on GitHub (Jan 28, 2026): Doesn't seem to happen anymore on 1.1.40 Thanks!
Author
Owner

@wangshuz commented on GitHub (Feb 5, 2026):

@inteist Will this occur on 1.1.50?

Thanks!

@wangshuz commented on GitHub (Feb 5, 2026): @inteist Will this occur on 1.1.50? Thanks!
Author
Owner

@ks-extlm commented on GitHub (Feb 5, 2026):

Nearly all issues related to excessive premium request usage have been addressed. Optimizations concerning the use of free, smaller models are detailed in the comments in #8030.

@ks-extlm commented on GitHub (Feb 5, 2026): Nearly all issues related to excessive premium request usage have been addressed. Optimizations concerning the use of free, smaller models are detailed in the comments in #8030.
Sign in to join this conversation.
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: anomalyco/opencode#7562