Today we're shipping an academic discount: 20% off any SPARKIT subscription for verified academic researchers, applied automatically at checkout when you sign up with an academic-domain email.
Plus drops to $40/mo, Pro to $80/mo, and Max to $240/mo. Same product, same query quotas, same overage rates — just cheaper for academics. The discount is lifetime, not a first-year promotion.
SPARKIT is a deep-research agent: send it a hard scientific question, get back a cited Markdown report. The natural home for that workflow is academic research — labs that read across hundreds of papers a week, postdocs scoping a new project, PIs evaluating evidence for a grant. We want SPARKIT in those workflows, and we don't want price to be why a curious researcher decides not to bother.
The 20% number is roughly where the math worked: meaningful enough to drive academic adoption, small enough to preserve margin on the most-utilized tiers, and clean enough that the discounted prices ($40 / $80 / $240) read as deliberate rather than awkward.
Any signup whose email host ends in one of these academic top-level domains:
.edu (US) · .ac.uk (UK) · .edu.au (AU) · .ac.jp (JP) · .edu.sg (SG) · .edu.cn (CN) · .ac.nz (NZ) · .ac.il (IL) · .edu.tw (TW) · .ac.kr (KR)
We didn't want to make this US-only — academic researchers outside the US are part of the user base we're building for. If your country's academic suffix isn't on this list and should be, send a note to info@sparkit.science and we'll add it.
That's the whole flow. No application, no certificate upload, no advisor approval, no annual reverification. We trust the email-domain check, and we'd rather under-verify and have a few alumni claim the discount than make every researcher prove their job to us every year.
The pricing page now has an Industry / Academia toggle. Switch to Academia and the discounted prices light up across all three subscription tiers, with strike-throughs on the original numbers so the savings are visible at a glance.
If you're at a lab thinking about a multi-seat plan, or a department wanting to license SPARKIT for a course or a cohort, the academic discount stacks on the regular Plus/Pro/Max tiers but isn't designed for institutional volume. For anything multi-seat or course-related, reach out at info@sparkit.science and we'll scope it directly — usually that ends up looking better than three Pro subscriptions.
Otherwise, sign up, and let us know what you build with it.
We ran a single GAIA question through SPARKIT, direct Claude Opus 4.7, and direct GPT-5.5. SPARKIT fetched Nature's archive, counted 1,002 articles, and answered correctly. Both direct LLMs invented different article counts and confidently landed on the wrong answer.
Read post →We ran the same hard scientific question through Perplexity, ChatGPT Deep Research, Gemini Deep Research, Elicit, and SPARKIT. All five got it right. Here is what was actually different — and why only one of them runs inside your codebase.
Read post →An end-to-end walkthrough using a real HLE-Gold question: install the SDK, send the call, read the report, and see the agent trace behind a 53% benchmark answer.
Read post →
SPARKIT