Data · June 2026
Measured, not marketed · synthesized in UltraDB

The 2026 Keyboards breakdown.

Mechanical and Hall-effect keyboards compared by what they actually do — switch feel, build, latency, connectivity and tunability — ranked by value. Wired input latency is independently measured (RTINGS / HL Planet); manufacturer claims are flagged.

Start with the job — it drives every chart below

Price vs capability

Each bubble is a keyboard; further right = more capable overall; color = brand/family. Re-scores when you pick a job — 'Competitive gaming' leans on latency + analog switches, 'Typing / enthusiast' leans on build + switch feel.

What works with what

Boards linked to the features they bring — Hall-effect analog, Rapid Trigger, SOCD, hot-swap, QMK/VIA, wireless and more.

Compatibility — keyboards ↔ switch & connectivity ecosystems

All keyboards

Method & honesty

Capability index — the formula (audit it)

Capability is the average of the five scored axes (latency, switch & typing, build, connectivity, features); value = capability² ÷ price. Responsiveness is scored from CLAIMED polling rate + switch technology, not independently measured latency (no comparable cross-brand measured-latency source).

The headline is an average — no board wins every axis. A Wooting 80HE tops gaming latency but a Keychron Q5 Max wins build & typing; pick the job chip that matches you. Measured latency beats spec-sheet polling-rate claims.

Honesty. Responsiveness uses manufacturer-claimed polling rate + switch tech — directional, not lab-measured latency (RTINGS-class measured data wasn't independently obtainable). Accuracy is priority #1 — nothing fabricated. Wired input latency is independently measured (HL Planet bench; RTINGS corroborates the latency leader); where only a polling-rate spec exists the ms is conservatively estimated and flagged. Prices are mid-2026 MSRP/street and move with sales. RTINGS' roundup pages bot-wall scrapers, so per-board RTINGS scores are summarized, not deep-scraped.