Why I Almost Switched to Raycast (And Why I Didn't Fully Commit)

Nearly a decade with Alfred. That's how long this relationship has lasted. Through three MacBook upgrades, two job changes, and countless macOS updates, Alfred remained the one constant in my workflow. Slack debates? Defended it. Custom workflows? Built dozens that saved me hours each week. Loyalist doesn't even begin to cover it.

Then Marcus, a colleague of mine, pulled up Raycast during a pair programming session. He searched for a GitHub PR, viewed the diff, and merged it without ever leaving the launcher. My first thought? "Okay, that's a neat trick." About thirty seconds later, a second thought crept in: "Wait, why can't Alfred do that?"

That moment kicked off six months of obsessive tracking and comparison. Every detail logged: time saved, friction points, features used, features missed. What surprised me was how much more nuanced the "which one's better" question turned out to be.

Here's the honest breakdown, complete with actual data and the workflows I rebuilt from scratch.

The Methodology: How I Tracked Productivity Metrics

Let's talk measurement, because "it feels faster" doesn't cut it when you're considering switching tools you use hundreds of times daily.

Timing.app tracked my application switches and time spent in each launcher. For the first two months, every launcher interaction went into a Notion database, noting what I searched for and whether I found it. Tedious? Absolutely. But after shipping products to millions of users at Google, I learned that gut feelings about productivity are usually wrong.

Here's how the experiment broke down:

• Months 1–2: Alfred only (my baseline)
• Months 3–4: Raycast only (cold turkey switch)
• Months 5–6: Both tools running simultaneously (the hybrid approach)

Five key metrics got tracked: launcher activations per day, average time from activation to task completion, failed searches requiring a different tool, new automations created, and subjective satisfaction scores at week's end.

The results? Not what I predicted.

Extension Ecosystems Compared: The Real Productivity Battlefield

When you're evaluating Raycast vs. Alfred for developers, the extension ecosystem determines whether you're using a fancy app launcher or a legitimate command center for your entire development workflow. Extensions make the difference.

GitHub Integration

Raycast's GitHub extension does exactly what you'd hope. Searching PRs, viewing details, interacting with repositories directly from the launcher. During month three, I measured this specifically: roughly 4–7 seconds saved per GitHub interaction compared to my Alfred workflow, which required opening the browser first. Multiply that by dozens of daily interactions, and it adds up fast.

Alfred's GitHub workflows exist, but they're community-maintained and vary wildly in quality. The best ones are powerful. The mediocre ones break with API changes and sit abandoned on GitHub repos.

Winner: Raycast, by a significant margin.

npm and Package Management

Here's where the comparison got interesting. Years ago, I wrote a custom Alfred workflow that searches npm packages and copies install commands. Building it took an afternoon, and it works exactly how I want it to.

Raycast's npm extension is good out of the box but less customizable. My specific workflow involves checking package sizes before installing, and Alfred's custom solution still wins for that use case.

Winner: Tie, depending on whether you value customization or convenience.

Docker Integration

Managing containers directly from the launcher? Raycast's Docker extension handles this beautifully. No more context-switching between apps just to restart a container. Doesn't sound like a big deal until you notice how often you're doing it.

Alfred requires custom workflows or third-party solutions that haven't seen updates in years.

Winner: Raycast, and it's not close.

Database Connections

Neither tool handles this particularly well out of the box. Custom script commands became necessary in both cases. The experience felt similar, though Raycast's TypeScript-based extension development came more naturally than Alfred's workflow approach.

Winner: Slight edge to Raycast for developer experience.

Developer-Specific Features: Where Each Tool Shines

Breaking down the best Raycast extensions for GitHub integration and other developer tools, compared to Alfred's equivalent features.

Script Commands

Script commands in Raycast work with Bash, Python, Ruby, or Swift. They integrate beautifully with the UI, supporting forms, dropdowns, and progress indicators. Building a script that spins up local development environments with environment variable selection took maybe two hours.

Alfred's equivalent offers more power but steeper learning. Various output formats in workflows give you control over how results appear, but you're working against more friction to get there.

Developers who want to automate repetitive coding tasks with Raycast on Mac will find a gentler learning curve. Those who want maximum control will find Alfred's ceiling higher. Pick your poison.

Clipboard History

Both tools offer clipboard history. Raycast's is prettier. Alfred's is more searchable and handles larger histories better. My clipboard keeps about 90 days of history (don't judge me), and Alfred handles the database more gracefully.

Small thing, but it matters when you're copying code snippets dozens of times daily.

Snippet Workflows

Snippets are central to my workflow: email templates, code boilerplate, Git commit formats, common terminal commands. Sound familiar?

Alfred's snippet system with auto-expansion has been rock solid for years. Type ;gitfix and it expands to my commit message format for bug fixes. Raycast's snippets work similarly but showed occasional expansion delays during testing.

For snippet-heavy developers: Alfred still edges out.

The Hidden Costs: ROI Analysis for Different Developer Salaries

Let's talk money, because whether Raycast Pro's subscription is worth the price depends entirely on your situation.

Alfred Powerpack: £34 one-time (single license) or £59 (Mega Supporter with lifetime upgrades). British pounds mean the USD equivalent varies with exchange rates.

Raycast Pro: $8/month when billed annually ($96/year), or $10/month when billed monthly. Pricing may change over time.

Over five years, that breaks down to:

• Alfred Mega Supporter: ~$75 total (at typical exchange rates)
• Raycast Pro: $480 total (at $96/year)

Significant difference. So is Raycast worth it for coding despite the higher cost?

Here's my framework. If Raycast saves you 10 minutes per day over Alfred, and your loaded cost to your employer is $100/hour, that's roughly $16 per workday, or about $340 per month. The subscription pays for itself in less than a day each month.

But here's the catch: during my experiment, Raycast saved me about 3–4 minutes daily over my optimized Alfred setup. New users who haven't invested years in Alfred customization would likely see higher savings.

My honest take: starting fresh? Raycast's productivity features make a strong case for the subscription. Already built extensive Alfred workflows? The switching cost plus subscription cost becomes harder to justify.

Migration Guide: What Transfers, What Breaks, What Disappears

For those considering the switch, here's what I learned about migrating between these tools. Some lessons came the hard way.

Transfers Easily

• Basic snippets: Export from Alfred, import to Raycast with minor reformatting
• Hotkey muscle memory: Both tools let you set custom activation shortcuts
• Basic workflows: Simple "search and open" workflows have equivalents

Requires Rebuilding

• Complex workflows with conditional logic: Alfred's workflow builder has no direct equivalent. You'll rewrite these as Raycast script commands or extensions.
• Third-party integrations: Many Alfred workflows connect to services in specific ways. Raycast extensions for the same services often work differently.
• Custom themes: Start over. They're not compatible.

You'll Lose (At Least as of My Testing)

• File actions and buffer: Alfred's file buffer system for batch operations didn't have a direct equivalent in Raycast during my testing period.
• Remote trigger: Alfred's remote iOS app didn't have a Raycast equivalent when I conducted this experiment, though this may have changed.
• 1Password integration depth: Alfred's 1Password workflow felt more mature than Raycast's extension during my testing.

Getting to "comfortable" took about six hours spread across a week. Reaching the point where I stopped instinctively grabbing for features that didn't exist? Another month. Plan accordingly.

The Verdict: Matching the Tool to the Developer

After six months of obsessive tracking, here's my honest breakdown of which developers should choose each tool.

Choose Raycast If:

• You're starting fresh with no launcher investment
• Your workflow centers heavily on GitHub, Linear, or Notion
• You prefer beautiful, polished interfaces
• Powerful features without building them yourself sounds appealing
• Subscription cost is negligible compared to your time value

Choose Alfred If:

• You've already built extensive custom workflows
• One-time purchases beat subscriptions in your book
• File manipulation is central to your workflow
• Maximum customization control matters to you
• You're price-sensitive and patient enough to build what you need

Consider Running Both If:

• You're a power user who can handle the cognitive overhead
• Specific features from each tool are non-negotiable
• You're still evaluating and want hands-on comparison time

Ended up in the third camp myself. Sounds ridiculous, I know. But here's the thing: Raycast handles my GitHub and API workflows better, while Alfred manages my snippets and file operations better. Different activation hotkeys keep the tools out of each other's way.

Is this the "right" answer? Probably not for most people. But it's my honest answer after six months of data.

The best Mac launcher app for programmers is whichever one removes friction from your specific workflow. Some developers thrive with Raycast's polish and integrations. Others prefer Alfred's flexibility and one-time cost. And weirdos like me who can't choose? Both works fine.

Start with whichever one sounds more appealing. Give it a real month. Then decide. The worst choice is spending more time debating than actually using either tool.

---

What launcher do you use, and what's your killer workflow? Always looking to steal good ideas. Drop a comment or find me on Mastodon.