How to Install Solver in Excel — Complete Guide (2026)

Most people who go looking for Solver assume it's a paid plugin. It isn't. Solver is a free optimization engine Microsoft bundles with Excel — and it's been there since Excel 3.0. The only reason you can't see it right now is that Microsoft ships it disabled by default. Flip one switch and the Solver button appears on the Data tab.

This guide walks you through the install on Windows, the install on Mac, the Excel Online limitation, and the first real model you'll build. You'll learn the three solving methods (GRG Nonlinear, Simplex LP, Evolutionary), when to use each one, and the fixes for the two errors that catch everyone — "Solver could not find a feasible solution" and the missing-button panic.

Quick context before we install. Solver was built by a company called Frontline Systems, which still develops the Premium versions today. Microsoft licenses the standard edition and ships it free inside Excel. That's why Solver lives under "Add-ins" rather than under regular Excel features — it's technically a third-party add-on that Microsoft preloads for you. The implication: turning it on is free, but it's a separate component, so it can also be turned off accidentally.

Install Solver on Windows (Excel 2016, 2019, 2021, 365)

Here's the exact click path. It takes about 20 seconds. Open Excel. Click File in the top-left ribbon. From the left sidebar, pick Options — you'll see a dialog box pop up. In that dialog, click Add-ins on the left.

Look at the bottom of the Add-ins panel. There's a dropdown labeled Manage: with "Excel Add-ins" already selected. Click the Go… button next to it. A small checklist window opens — that's where the magic happens. Tick the box next to Solver Add-in. Click OK. Done. No restart, no download.

Now click the Data tab on the main ribbon. Scan all the way to the right. You'll see a new group called Analyze with a Solver button inside it. That's your install confirmation. If you also see Data Analysis next to it, that's the Analysis ToolPak — same install path, different checkbox, useful for regression and ANOVA.

Worth knowing: Solver settings are per-workbook, not per-install. If you email a workbook with Solver constraints set up, the recipient sees those constraints — but only if they've enabled Solver on their copy of Excel too. The model travels. The add-in doesn't. That's the most common cause of "my colleague's file doesn't work on my Excel" complaints.

One more Windows detail: if your IT department locked down Add-ins (common in finance and healthcare environments), the Solver checkbox may be greyed out. Check with admin — usually a one-line policy change unblocks it. Solver itself doesn't touch the network, file system, or registry, so most IT teams approve it without fuss once asked.

Install Solver on Mac (Excel for Microsoft 365, 2019, 2021)

Mac users sometimes can't find the Add-ins panel because Microsoft put it somewhere different. Don't go hunting in Preferences — it's in the menu bar. Open Excel. From the top menu, click Tools. From the dropdown, pick Excel Add-ins…. The same checklist window from Windows pops up. Tick Solver. Click OK.

If you're on Excel 2011 for Mac, the path is slightly different — Solver lives under Tools but as its own entry, and you may need to download it separately from Microsoft. Excel 2016 onward fixed that. Anything from the last decade follows the click path above.

After enabling, head to the Data tab on the ribbon. Solver appears on the far right. That's it. The Mac and Windows versions of Solver share the same engine, so any constraints, objective cells, and changing cells you set up on one platform will run identically on the other. You can hand a Solver workbook to a Windows colleague — or get one from them — without conversion headaches.

One Mac gotcha: if you have multiple Office installs (a 2019 standalone plus a 365 install, say), make sure you're enabling Solver on the version you actually use. The checkbox is per-application, not per-Mac. Quitting one Excel and opening another won't carry the setting over. Worth double-checking which Excel is your default if Solver "disappears" between sessions.

Apple Silicon Macs (M1, M2, M3) run Solver natively — there's no Rosetta translation involved as of Excel for Mac 2021 and later. Speed is comparable to a similarly priced Intel machine, sometimes faster on big models because the M-series chips handle the Solver iteration loop efficiently. Older Intel Macs still work fine for everyday models too.

Excel Online — No Solver, Here's the Workaround

This catches people off guard. Excel Online (the version you open at office.com or through SharePoint) does not include Solver. It's not a hidden toggle, it's just not built. Same goes for Excel for iPad and Excel for iPhone — Solver is desktop-only.

The workaround is simple. If a workbook with Solver lands in your OneDrive and you want to actually run it, click Open in Desktop App from the Excel Online toolbar. The file opens in your local Excel, where Solver works normally. Save changes, and they sync back to OneDrive.

No desktop Excel at all? You've got two free options. LibreOffice Calc includes its own Solver with similar capabilities — the UI differs but the math is the same. Download from libreoffice.org. Or try Google Sheets with the OpenSolver add-on from the Google Workspace Marketplace, which runs linear and integer problems in your browser. Neither matches desktop Excel for huge models, but for everyday optimization both work.

Running Your First Solver Model

Solver's interface looks intimidating the first time you open it. It isn't. Every Solver model has exactly three things: an Objective (the cell you want to maximize, minimize, or hit a target on), Variable Cells (the cells Solver is allowed to change), and Constraints (the rules Solver has to respect). That's it.

Quick example. Imagine you sell two products — A profits $5 each, B profits $8 each. You have 40 hours of labor available. A takes 2 hours per unit, B takes 4 hours. Maximum demand is 15 units of A and 10 units of B. How many of each should you make?

Set up a sheet. Put units of A in B1 and units of B in B2 (start with 0). Profit formula in B3: =B1*5+B2*8. Labor used in B4: =B1*2+B2*4. Now click Data → Solver. Set Objective = B3. Click "Max". By Changing Variable Cells = B1:B2. Add constraints: B4 ≤ 40, B1 ≤ 15, B2 ≤ 10, B1:B2 ≥ 0, B1:B2 = integer. Solving Method = Simplex LP (linear problem). Click Solve.

You'll see the answer in about half a second: 10 units of B, 0 of A, profit $80, labor maxed at 40 hours. Solver figured out that B's higher margin beats A's lower labor cost. That's the entire workflow, and it scales the same way to 1,000-variable problems.

Worth knowing for bigger models: tick Make Unconstrained Variables Non-Negative in the main Solver dialog unless you genuinely allow negative values. Most real-world problems (units produced, dollars invested, hours worked) can't go negative, and forgetting this constraint leads to nonsense results that Solver will happily report as "optimal."

After Solver finishes, you'll see a Results dialog. Click Answer Report in the list on the right before clicking OK. It generates a new worksheet with the final values, constraint status (binding vs not binding), and slack for each rule. The Sensitivity Report next to it tells you which constraints, if relaxed by one unit, would improve your objective the most — invaluable for follow-up analysis.

The third option, Limits Report, shows the upper and lower bound each variable could take while keeping the solution feasible. Most beginners skip these reports and miss the actual insight Solver provides — the solution itself is half the value, the reports are the other half.

Choosing the Right Solving Method

Solver gives you three engines and they aren't interchangeable. Pick the wrong one and you'll either get a slow answer, a wrong answer, or no answer at all. Here's the honest breakdown.

Simplex LP is for linear problems — anything where doubling an input doubles the output, and there are no curves, squares, or IF statements in the formula chain. Classic use cases: product mix, blending, transportation, scheduling. Simplex is the fastest of the three by a wide margin. If your model is truly linear, Solver will find the global optimum in milliseconds, even with thousands of variables.

GRG Nonlinear (Generalized Reduced Gradient) is Solver's default. Use it when formulas have curves — quadratics, exponentials, logarithms — but stay smooth (no jagged jumps). Portfolio optimization with covariance matrices, regression curve fitting, and pricing models all fit here. GRG finds a local optimum, not a global one, which means starting values matter. Try a few different starting points if results look off.

Evolutionary is the slow but flexible option. Use it for non-smooth problems — formulas containing IF, VLOOKUP, MIN, MAX, ABS, or any logic that creates jumps in the solution surface. Examples: scheduling with shift constraints, knapsack-style packing, anything where Excel's calculation jumps suddenly. Evolutionary uses genetic algorithms, so it tries lots of random combinations. Expect 30 seconds to several minutes on medium models. It's an approximation — it won't guarantee the absolute best answer, just a good one.

Quick test: if your formulas only use +, -, *, /, SUM, SUMPRODUCT — Simplex LP. If they use ^, EXP, LOG, sin/cos — GRG. If they use IF, VLOOKUP, MAX/MIN, ABS — Evolutionary. When in doubt, try Simplex first; if it complains about non-linearity, switch to GRG. If GRG returns weird results, fall back to Evolutionary.

One more tip on method selection. Solver remembers your last choice per workbook, so if you opened someone else's file you might be running on whatever engine they last picked. Always glance at the Solving Method dropdown before clicking Solve — a Simplex setting on a nonlinear model will silently produce wrong answers without raising an error. Triple-check on inherited workbooks.

What People Actually Use Solver For

Solver isn't a toy. It runs production-scale optimization at thousands of companies. The common use cases below are the ones where it shines on a standard Excel install — no Premium Solver Platform needed.

Real-World Examples Where Solver Earns Its Keep

Linear optimization is the everyday workhorse. Manufacturing companies use it to decide how many of each product to make given limited machine hours and raw materials. Schedulers use it to assign nurses to shifts while respecting overtime caps. Logistics teams use it to pick which warehouse ships to which customer at lowest total cost. All of these problems share a structure: a clear objective, finite resources, and rules that can be written as linear formulas.

Portfolio optimization is where GRG Nonlinear earns its place. Given a list of assets with expected returns and a covariance matrix of how they move together, Solver finds the mix that maximizes return for a given risk level — Markowitz's classic mean-variance frontier, built in Excel in 20 minutes. It's not Bloomberg, but for personal portfolios or small fund work, it's plenty.

Regression curve fitting uses Solver to find the best parameters for any custom equation you can write — useful when your data doesn't fit Excel's built-in LINEST or TREND functions. Set up your model with parameters as variable cells, minimize the sum of squared errors, and Solver hands you the best-fit values. Common in scientific work where you need a specific non-linear curve shape.

Schedule and assignment problems are where Evolutionary shines. Assigning students, employees, or deliveries — anywhere you have yes/no decisions plus rules that can't easily be written linearly. Slow but flexible, and it usually beats hand-tuned solutions by 10-30% on real datasets. The excel pivot tables guide is the next step once your optimization output needs slicing.

Troubleshooting Common Solver Errors

"Solver Add-in is not installed" appears most often after a Microsoft 365 update wipes your settings. Fix: redo the install path from the top of this guide. File → Options → Add-ins → Manage: Excel Add-ins → Go → tick Solver. Two minutes max.

"Solver could not find a feasible solution" means your constraints contradict each other. Example: you require total production ≥ 100 units but also total labor ≤ 20 hours when each unit needs 1 hour. There's no way to satisfy both. Fix: review constraints one at a time, comment out the most restrictive ones, and re-run to find the conflict. The Sensitivity Report Solver offers after a successful run will flag which constraints are binding.

"The Objective Cell values do not converge" means Solver thinks the answer is infinite — usually because you forgot an upper-bound constraint. If you tell Solver to maximize profit with no cap on production, it will happily say "make a billion units" and then complain. Add a realistic upper bound (capacity, demand) and re-run.

One sneaky cause of the missing-button panic: Excel sometimes disables Solver after a crash or unexpected shutdown to protect the workbook. Check File → Options → Add-ins and look at the Disabled Application Add-ins section near the bottom. If Solver shows up there, click Manage: Disabled Items → Go → enable Solver, then re-enable it in the regular Add-ins panel. Two-step fix, three minutes.

If you're working with text-heavy reference sheets while building your model, the how to indent in excel walkthrough is worth bookmarking — clean indentation makes constraint lists readable when your model grows past 20 variables. And when you're ready to test your Excel skills broadly, the excel hub has the full library of practice tests across formulas, functions, and data analysis tools. Worth the bookmark before your next interview prep round.