Token Beggar 0.3.0: on 18 real bug fixes it cost 1.42x less, with a tie on tests

Description and use

Token Beggar 0.3.0 cuts bug‑fix costs by 1.42× while keeping resolution rates similar, though it does not speed up execution.

Problem

Running the largest model for every step of a code‑fix task is expensive; a cheaper model could handle many tasks without sacrificing quality.

Outcome

Token Beggar 0.3.0 lowered the cost per bug‑fix task by 1.42× (from $0.69 to $0.49) across 18 hard tasks, with issue resolution remaining at 69% versus 70% and runtime unchanged.

Workflow

• 1. Plugin intercepts a task request.
• 2. It estimates task size based on repository context.
• 3. Small tasks are handled by the chat model directly.
• 4. Large tasks are sent to a cheaper worker model for reproduction, exploration, fixing, and test execution.
• 5. The worker returns a diff and test results.
• 6. The chat model reviews the diff, re‑runs tests, and if needed, sends the task back to the worker for further refinement.
• 7. Final diff is applied to the repository.

Evidence

Without Token Beggar is the chat model working alone. With Token Beggar is the card shipped in 0.3.0.

Source

Cost per task went down about 29 percent.

Source

The share of issues resolved is level, and 69% against 70% is a tie within the spread.

Source

Time is level too (168 seconds against 187), so it is not faster.

Source

The raw runs and the task list are in the repo, so you can check every number.

Source