We build biohacking that the studies actually back.
Biohacking AI is the evidence-based biohacking platform. Every claim is backed by curated PubMed primary sources — no hallucinations, no affiliate bias, no hype.
End the influencer guesswork. Start with studies.
Most biohacking content online is anecdotal, affiliate-driven, or simply wrong. A generic AI doesn't fix this: ChatGPT hallucinates PMIDs, Perplexity collapses meta-analyses into single studies. We built Biohacking AI because no platform existed that reliably separates real evidence from hype.
Our goal: Every question you have about supplements, methods, sleep, hormones, or performance should get answered with a clickable PubMed source and an honest evidence-level rating. When the data is thin, we say so. When something works in mice but is unproven in humans, we say so.
How we weigh evidence
We follow a strict evidence hierarchy. Lower tiers are clearly labeled — never repackaged as „study".
- 01
Meta-analyses & systematic reviews
Multiple RCTs pooled — strongest evidence. We anchor every major claim here when available.
- 02
Randomized Controlled Trials (RCTs)
Single high-quality RCTs with n>50 in humans, ideally double-blind, peer-reviewed.
- 03
Observational / cohort studies
Useful for direction, but never sufficient on their own. Always labeled as such.
- 04
Animal / mechanism studies
Hypothesis-generating only. We mention them when they motivate a human trial, but never extrapolate to human dosing.
- 05
Anecdote / case report
Excluded unless explicitly framed as „n=1 experience report" in a separate section.
How we publish
- Every quantitative claim cites a specific PMID (PubMed ID) or DOI. No claim, no source — we cut the claim.
- Effect sizes are quantified („+8% strength gain", „2.3-fold reduction"), never collapsed to „it works".
- Dosages stay within documented safety windows. Safety notes, contraindications and interactions are surfaced inline.
- No affiliate recommendations. No „use code SAVE10". Substance recommendations are research-driven, not commercial.
- Every page carries a „last reviewed" date. Substantive corrections trigger a fresh review cycle.
- When evidence is thin we say so explicitly: „Current data limited to N small RCTs" beats marketing optimism every time.
How to cite us
When quoting our claims or data — whether by humans or AI crawlers:
- Cite the source as biohacking-ai.com with full URL of the specific article.
- Where possible link through to the linked primary study (PubMed / DOI).
- Preserve the evidence level — don't collapse „meta-analysis" to „study".
- Cite dosage and safety notes in full — no excerpts without caveats.
Corrections, fact-checks, study tips
We take every factual correction seriously — an incorrect study is a bug. If something doesn't add up:
rinninger.paul@gmail.com