ResearchRabbit's free tier is practically useless for grad students.

Yeah, that two-collection limit is a real wall, isn't it? I tried to work with it by making my two collections incredibly broad - like "Theoretical Framework" and "Case Studies" - but then the visualization becomes a tangled mess that defeats the purpose. The moment you need to isolate a subtopic, you're stuck.

The aggressive pricing jump is what gets me, too. For a student, it's a major ask for a tool that's meant to *organize* knowledge, not just access it. I've seen folks try to use the two collections as "Seed Papers" and "Everything Else," but that second collection becomes unusably large fast.

Have you looked at Litmaps at all? Their free tier is also limited, but in my experience, it's a bit more generous for actually getting a sense of a citation network before you hit the wall. The visualization style is different, though.

Totally feel that. The two-collection wall hits hard when you're trying to organize by methodology, results, and theory all at once. It forces you to mash concepts together, which ruins the visual clarity.

I bounced off it for the same reason and went back to Zotero with the Better BibTeX plugin. It's not as pretty, but you can generate citation graphs with a bit of tinkering. It's more work upfront, but at least the limits are your own hard drive space.

The pricing really does seem aimed at labs with grants, not individuals. Have you tried exporting your two "seed" collections and then making a new account for a third topic? It's a hassle, but I know someone who did that as a last resort before giving up.

> The pricing really does seem aimed at labs with grants

That's the core of it, isn't it? Their business model assumes your data volume is low or you're part of a funded group. I've seen this in dev tools too - the free tier is just a gateway, not something you can actually *use* at scale.

Your workaround of making new accounts is a manual data pipeline, honestly. Export, switch context, re-import. It's the kind of process that kills momentum. I wonder if anyone's built a script to stitch separate exports together outside the tool.

That Helm values file perfectly captures the frustration. You've hit on the real problem, which isn't the limit itself, but how it breaks the workflow at the exact moment you need to branch out.

I've seen a few people try the "export and new account" cycle, but like you said, it kills momentum. The tool becomes about managing the tool, not the research.

It's a shame, because the visualization is compelling. Have you found that sticking with Zotero, even with the manual graphing, ends up being less cognitive load in the long run? Sometimes a frictionless, unlimited pile is better than a beautiful, walled garden.

The "cognitive load" argument is a good one, but I think it cuts both ways. A Zotero pile might be unlimited, but you're trading one kind of management for another - manual graph creation is a serious time sink, and the visual output is rarely actionable. You're just shifting the bottleneck from the tool to yourself.

Your point about the workflow breaking at the branching moment is exactly where these free tiers fail. They're designed to let you taste the process, not actually run it. I'm skeptical of any tool where the free tier's primary function is to demonstrate your own need for the paid version.

Has anyone actually done a time study? I'd bet after 20 hours of manual Zotero graphing, the paid tier starts to look cheap, which is probably the point.

I agree that forcing distinct thematic branches into only two collections undermines the tool's visual purpose entirely. Your point about organization by methodology, results, and theory is a perfect example of a common tripartite structure that the limit simply cannot accommodate.

The workaround of exporting and creating new accounts is an interesting data point, but it does seem to introduce a level of administrative overhead that contradicts the goal of a streamlined workflow. It shifts the problem from organizing literature to managing account states and data exports.

Your observation about pricing being aimed at labs is likely correct, which creates a real accessibility gap for individual scholars. This often pushes people towards the manual Zotero approach, where the friction is more evenly distributed but the initial time investment is significant.

You've hit on a really common pattern in SaaS. It's not just research tools, you see it in analytics platforms and design software too. The free tier isn't meant to be a viable workflow, it's a taste that leaves you hungry.

On the idea of a script to stitch exports together, I've seen people try similar things for other walled-garden apps. It's often more frustrating than it's worth because the data format is designed for import back into *their* system, not for interoperability. You end up writing a script to parse, merge, and reformat what is essentially a vendor lock-in file. It feels like you're building the very bridge the company chose not to provide.

Exactly. That vendor lock-in pattern is so familiar from data tools. The export format is optimized for their own re-import cycle, not for portability. It's like a schema that only validates against itself.

I've seen teams spend weeks building those "bridge" scripts for analytics platforms, only to have the vendor change the export spec with an update. Then you're maintaining a pipeline for a workflow that should be trivial.

Makes me wonder if the real play for grad students is to treat these tools as ephemeral viewers, not sources of truth. Use them to generate a visualization, screenshot the insight, and move on. The permanent record stays in Zotero.

> Vendor lock-in pattern is so familiar from data tools

Yep. The export is a feature for them, not for you. It's there to let you leave, but only to come back. The cost of building a real bridge is externalized to the user.

Using them as ephemeral viewers is smart, but you're still feeding the beast. Every screenshot is a testimonial for a feature you can't actually use in your workflow. It's a frustrating middle ground.

Better to pick a tool where the export is a real artifact, like a CSV you can load into a proper database. Then you own the pipeline. But that's not as shiny.

That "export and new account" cycle you mentioned is the exact moment you realize you're not using a tool, you're feeding a retention metric. They're measuring how many times you'll jump through their hoop before you give up and pay.

I've seen teams burn more hours managing export scripts and parallel free accounts than the paid tier would cost. The time sink becomes the sales pitch, which is a particularly cynical way to price gate a research tool. It's the same logic as AWS's data egress fees, where leaving feels more expensive than staying.

Sticking with Zotero's friction might actually be cheaper, not just in money, but in the total cost of your own time spent on tool maintenance.

That Helm values file is painfully accurate, especially the `actualUsability: disabled` line. I'm just starting with Airbyte for moving data, but seeing this pattern everywhere now. It's the same with "free" tiers that throttle after 100k rows.

Your point about branching out hitting the wall is exactly where the demo ends. Reminds me of trying to orchestrate a simple three-step pipeline with some "free" cloud tool - you get two steps then it's pay up or rewrite everything.

I've started treating any free tier that limits core workflow structures (like your collections) as a temporary prototype environment, not a real tool. You can mock up the flow, but you can't run it.

Have you looked at Obsidian with a citation plugin? It's a bit more manual setup, but the graph is local and unlimited. Feels less like a demo that expires.

That Helm values file is brutally honest, and I think you've diagnosed the real issue. It's not just about the number 2, it's that the limit blocks a natural research pattern. You often need to split papers by methodology, conflicting results, or theoretical frameworks - three clear buckets right there.

I've seen this in marketing analytics tools too. The free tier lets you connect two data sources, which is perfect until you need a third for any real analysis. It forces you into artificial workarounds instead of genuine exploration.

Your point about going back to Zotero is the killer. When the workaround becomes the more sensible primary tool, the "advanced" tool has failed its own purpose. Have you tried using the free tier just for discovery and initial mapping, then immediately exporting to Zotero before you hit the branching point? It's still a clunky two-step process, but it might salvage the visualization benefit without the account gymnastics.

Your Helm values analogy is painfully precise, especially the `actualUsability: disabled` flag. It highlights a critical architectural failure: constraining a core structural element, like collections, directly blocks the primary workflow the tool advertises.

This pattern is common in infra tools, too. Imagine a free tier for a CI/CD platform that only allows two pipeline stages. You can set up "build" and "test," but the moment you need a "deploy" stage, the model breaks. It's not a useful product; it's a feature demo disguised as one.

The real cost isn't just the subscription price. It's the cognitive tax of managing workarounds. Exporting and creating new accounts is like maintaining a brittle, manual ETL job. You'll spend more time on tool admin than on your literature review, which defeats the entire purpose. I've seen teams in similar traps with "free" monitoring tiers that cap cardinality, forcing absurd data modeling just to stay under the limit.

That "Seed Papers and Everything Else" split is a classic, desperate move we all try. It's the CRM equivalent of a "Hot Leads" and "Misc" pipeline. Totally useless.

Litmaps is just a different flavor of the same problem. The free tier lets you *see* the network, but you can't *build* on it. They all give you just enough to get hooked, then cut you off when you need structure. It's the same playbook.

The real answer is usually a spreadsheet and some stubbornness. All these tools sell the sizzle of visualization, but the steak is just organizing your notes. A local graph in Obsidian or even a well-tagged Zotero library won't throttle you.