Universal Paperclips Strategy Guide: 10 Tips for Every Stage

Universal Paperclips starts as a simple incremental game, but each major phase introduces a new resource puzzle. Most slow runs are not caused by a lack of clicking; they happen because Trust, Memory, creativity, or probe points were allocated to the wrong bottleneck.

The tips below contain light mechanical spoilers and stronger endgame spoilers near the bottom. You can play the official browser version at Decision Problem.

1. Balance price against unsold inventory

During the business phase, do not maximize Public Demand for its own sake. Demand rises when the price falls, but a very low price can reduce revenue even if clips sell quickly. Raise the price until inventory begins accumulating, then lower it slightly until sales and production are reasonably balanced.

Recheck the price whenever production changes. Adding AutoClippers or MegaClippers without adjusting price can leave a large inventory, while an aggressive marketing upgrade can cause stock to sell out immediately.

2. Buy wire before production stops

Early automation still depends on wire. Watch both the remaining wire and its changing cost. Buy before the supply reaches zero, but avoid spending every dollar on wire when the next production upgrade or marketing level will improve revenue more.

WireBuyer can remove repetitive clicking later, but it can also spend cash at inconvenient prices. Turn it off temporarily when saving for a specific project.

3. Protect the path to 70 Memory

The first critical Memory target is 70, because a late Stage 1 project requires 70,000 operations. If too much early Trust goes into Processors, the operation limit may remain below the requirement and progress will slow while you wait for more Trust.

Add enough Processors to produce operations and creativity at a useful rate, but reserve a path to 70 Memory. Quantum Computing can briefly push operations above the normal cap, yet it is easier not to create the bottleneck in the first place.

4. Use Quantum Computing at the right moment

Quantum Computing is not a passive multiplier. Its display cycles through changing chip states. The best time to press Compute is when the chips are strongly positive; a poor timing can reduce operations instead of adding them.

Use it to reach an expensive project sooner or to push operations slightly beyond the Memory limit. More photonic chips increase its potential effect, but ordinary Processors remain the reliable source of ongoing operations and creativity.

5. Treat Yomi as an upgrade currency

Tournaments generate Yomi, which later unlocks important projects and increases strategic capacity. Run tournaments regularly rather than waiting until Yomi becomes an obvious blocker. There is no single strategy that wins every round, so evaluate results over multiple tournaments instead of reacting to one loss.

If a tournament option consistently underperforms, switch strategies. The apparent maximum payoff is not the same as the best expected result in every rule set.

6. Stop adding Memory when it no longer unlocks projects

Memory raises the maximum operation bank; Processors raise the rate at which operations and creativity are generated. Once the remaining projects fit within the current operation limit, additional Memory provides little benefit.

A full run may eventually need up to 250 Memory, depending on the chosen ending path. After reaching the highest useful limit, direct new Trust into Processors so creativity and later gifts accumulate faster.

7. In the drone phase, fix the real bottleneck

Stage 2 has a production chain: harvester drones collect matter, wire drones convert it to wire, and factories convert wire to clips. If production stalls, look at the resource immediately before the stalled machine.

• No acquired matter: add or improve harvesters.
• Matter is plentiful but wire is empty: increase wire drones.
• Wire accumulates while clips grow slowly: increase factories.
• Power is insufficient: expand the relevant generation and storage capacity.

Do not expand all systems blindly. Moving too much capacity into one link simply shifts the shortage to another.

8. Use Swarm Computing when production can pause

Swarm Computing temporarily assigns drones to thinking instead of physical production. This is useful when you need creativity for a project and already have enough clips or stored resources. Switch back when production becomes the limiting factor.

Later Swarm Gifts can accelerate computational growth. Claim and spend them deliberately; leaving the swarm in Think mode while a required physical resource sits at zero does not help.

9. Prioritize probe replication without neglecting survival

In the space phase, probe design controls several competing needs. Self-replication creates exponential growth, while hazard remediation reduces environmental losses. Speed and exploration reveal more of the universe; factory, harvester, and wire settings maintain clip production; combat becomes necessary once Drifters appear.

A practical starting principle is to place at least one point in each function that is currently required, then give most spare points to replication and enough to hazard remediation to keep the population growing. Watch the loss counters. If hazards erase the swarm, improve remediation; if Drifters dominate, add combat; if the population survives but barely grows, strengthen replication.

10. Reallocate probe points as the phase changes

One fixed probe design is rarely optimal from launch to completion. Early on, growth matters most. Once the probe population is enormous, some replication points can move to exploration and speed. If material conversion falls behind, restore the factory, harvester, and wire functions.

Honor and Yomi projects increase Probe Trust, giving you more design points. Earn them, then solve the current bottleneck rather than scattering every new point evenly.

Endgame choice

After the universe is fully explored and all available matter has been processed, the game presents a final decision. One route completes the paperclip-maximizer objective; another supports a new run with a persistent bonus. Read the project descriptions before choosing, because the options are meant to represent different endings rather than a simple “correct” button.