← Decision ReportsUpdated 26 Jul 2026
Which general AI assistant should I use?
Tell us what matters most, and we'll tell you which one actually wins for that — including on the rare occasion where the honest answer is that nothing does yet.
What matters most to you?
Recommendation
—
—
MATCH SCORE
Adjust the weighting yourself
What we do know
- +Claude leads on getting things right, coding, research, long documents, writing quality and privacy protection — a broad, consistent lead across most of what we've checked
- +Gemini is clearly the fastest and the cheapest to run per task
- +Perplexity is the clear best choice specifically for citing trustworthy web sources
- +Perplexity and ChatGPT connect with far more outside apps and business tools than Claude or Gemini do
What we still don't know
- –How well each tool generates images
- –The exact coding numbers are contested — sources disagree on the precise gap, only on the ranking
What we checked, and how
| What we checked | How we know | Status |
|---|
We'll update this when
- We measure image generation — the one gap left
- The coding numbers get confirmed by one authoritative source instead of several that disagree
- Writing quality gets backed by an actual test instead of reviewer consensus
- Any new release changes the gap enough to flip an answer above
Price is estimated, not measured directly. We used each company's published API pricing and
assumed a typical task uses about 2,000 words of input and 1,000 words of output — a real number, but
a modeled one, not a bill we actually paid. A different kind of task could change the price ranking.