How I find OSRS botting money makers in 2026
How I research OSRS botting methods: script libraries, live bots, wiki trade volume, manual tests and the numbers I record before scaling a farm.
The useful answer to "what should I bot?" is a process, not an item name. By the time a public method is being pasted into every Discord, the margin is normally worse and the same script is running on hundreds of accounts.
I look for methods by starting with what scripts can run, checking whether the output will sell and testing the route myself. Only then do I put accounts on it.
Play enough OSRS to understand the method
If you have never played the game, do that first. You need to know why an account is walking to a bank, which requirement blocks the route and whether a wiki GP/hr number assumes attention your script cannot reproduce.
This is not me saying you need a max cape before opening DreamBot. Thirty minutes doing the proposed method by hand is often enough. You will notice the bank distance, dialogue, item limits and failure points that a spreadsheet misses.
It also helps you spot nonsense. A method needing ten quest unlocks and an expensive spec weapon is not a fresh-account farm just because somebody made a script for it.
Start with script libraries
Browse the SDNs for DreamBot, OSBot and whichever other clients you actually use. The public scripts show what has enough demand for somebody to maintain. Forum "Local Scripts" sections can be useful too. Those scripts are JARs or source releases that have not necessarily gone through the public SDN review, so read the thread and run them on an account you can lose.
Do not buy the first premium script with a good progress screenshot. Check the support thread. Search for recent reports of broken banking, missing areas or an update that has not been fixed. A pretty 100-hour proggy from last year does not tell you what the current build does.
Use trials where they exist. I want to see the actual GP/hr, CPU use and how the script recovers before I buy more auths.
Watch bots already in game
OSRS shows you live tests all day. If I see an account with 50m XP in one skill and almost nothing elsewhere, I check the activity and build. If another bot at the same place has several skills in the 50s and four times the XP, that difference is worth investigating.
RuneLite's hiscore lookup makes the first pass quick. Account age, quests and off-skill levels are not visible in full, so do not invent a ban theory from one account. I use the observation to make a test: low TTL versus developed, different route or different session length.
This is how a lot of my odd collecting projects started. Rakes, spades and watering cans were not secret high-profit methods. They were interesting because collectors would buy absurd quantities and I wanted to see how far the stockpile could go.
Use the wiki, then check trade volume
The wiki money making guide is the easiest list to work from. Filter by requirements and method type, then open the output item and check its live GE volume.
GP/hr without demand is useless. The old version of this post used buying beer tankards as the example: the calculated profit looked decent, but only around 5,000 were trading per day at the time. Two hours of output could equal the whole displayed daily market. The exact numbers have changed since 2025; the reason for checking them has not.
High-volume items such as common bars are easier to unload. An obscure item may show a much better margin and still leave every account holding stock for a week. Increased supply can create some extra volume, but do not assume the GE will absorb whatever your farm makes.
For less obvious ideas, I also check Unusual Money Making. For combat methods, the wiki bestiary lets you work from combat level, area and drops instead of beginning with the same five farmed monsters everyone knows.
Score the method before scaling it
I write down the same numbers for each test. Nothing fancy.
| Check | What I record | Why it can kill the method |
|---|---|---|
| Requirements | Stats, quests, gear and travel unlocks | Replacement accounts may cost more than the output |
| Real output | Items or GP per hour from my script config | Advertised rates often assume perfect conditions |
| GE demand | Daily volume and spread | Profit does not count until the stock sells |
| Stability | Stucks, deaths and manual fixes per day | A method needing babysitting does not scale |
| Account life | Hours run before a ban or lock | Gross GP means little without replacement cost |
| Infrastructure | RAM, CPU, proxies and membership | A lower-GP method can win if it is cheap to run |
Run the method manually first. Then run one to three disposable accounts long enough to expose route problems. I normally want more than one result because a single ban proves very little.
Work out net output after account, bond, proxy and script costs. A bot showing 500k/hr is not making that for you if it needs a $20 account every day.
Scale the boring winner
Once a test works, increase in steps you can still observe. Five accounts can show a shared banking bug that one account never hit. Twenty can show that the market was too small. Going straight to 100 makes it harder to tell which part failed.
I keep a few ordinary methods running while testing stranger ones. The ordinary farm pays the bills. Experiments are allowed to fail without taking the whole setup with them.
Methods also decay. Prices move, scripts break and ban reports cluster around activities. Keep the test sheet so you can see when GP per account falls below the replacement cost instead of continuing because the method worked last month.
The beginner botting guide covers test batches, updates and script selection. If account development is part of your cost calculation, read the TTL guide before treating every login as interchangeable. The builds I currently have available are listed on the stock page.