Fair Use Policy

Every job that reaches your UpHunt feed is scored by an AI model against your feed prompts. That scoring runs continuously, around the clock, for every feed you create, and each scored job costs real compute. For almost everyone this is invisible: normal use, even heavy normal use, sits comfortably inside the allowances below.

The allowances exist to keep the product sustainable when an account runs far outside normal use, for example scripted bulk feed creation or reselling scored output. They are set as ceilings on extreme consumption, not as meters to think about day to day.

Each plan includes a monthly AI scoring allowance, measured in model tokens. One scored job uses about 700 tokens on average, so the allowances translate to approximately:

• Free Trial: 20 million tokens for the 7-day trial. That is enough for roughly 28,000 scored jobs, or roughly 70 generated proposals, or any mix of the two.
• Basic: 200 million tokens per billing month. That is enough for roughly 280,000 scored jobs, or roughly 700 generated proposals, or any mix of the two.
• Auto-Apply: 480 million tokens per billing month. That is enough for roughly 680,000 scored jobs, or roughly 1,700 generated proposals, or any mix of the two.

The allowance covers all AI features on your account: job scoring, proposal generation, proposal refinement, and screening question answers. Job scoring is by far the largest share for a typical account.

Different features run on different AI models with very different costs, so usage is metered at cost-equivalent rates: a token spent on a premium proposal model (such as Claude Opus) draws proportionally more of your allowance than a job-scoring token. In practice, a single generated proposal uses roughly the same allowance as a few hundred scored jobs, and a typical account generating dozens of proposals a month stays far inside every allowance above.

Feed pre-filters do not consume the allowance. Keyword includes and excludes, minimum rates, and client filters (spend, rating, country) run before the AI, so tighter feeds make an allowance go a long way: only jobs that pass your filters are ever scored.

Your allowance period is anchored to the date you signed up and resets on that day each month, not on the calendar month.

If you reach your allowance before the period ends:

• Automatic job scoring pauses. New jobs are not scored or matched against your feeds until the period resets.
• AI generation pauses too. Requests to generate or refine a proposal, or to answer screening questions, return a clear allowance message with the reset date instead of running. Applications already in progress are completed normally.
• We tell you. You get a notification on your connected channels and a banner in your dashboard showing the exact date scoring resumes.
• Nothing is deleted and nothing else changes. Your feeds, settings, and existing matches stay exactly as they are, and scoring resumes automatically when the new period starts.
• Upgrading takes effect immediately. Moving to a higher plan raises your allowance at once, and scoring resumes right away if the new allowance covers your usage.

Some legitimate accounts, typically agencies running many distinct feeds, may need more than the standard Auto-Apply allowance. Contact us and we will look at your account; we can raise individual allowances where the usage is genuine. We would rather adjust your limit than have you work around it.

Accounts that use automation to extract scored output at scale, for example creating feeds by script in bulk or piping results into a competing service, are outside fair use regardless of allowance, and we may limit or close them under our terms. The allowances above are generous precisely so that no genuine freelancer or agency ever hits them by accident.

If we change the allowances, we will update this page and the change will apply from your next billing period, never retroactively within a period that has already started. This policy was introduced on 20 August 2026.