Beginner cost
Free tier, common paid plan about US$20/month, higher Max and Team plans.
Score working
Last updated: 3 June 2026
Anthropic
Claude is strong for writing, long documents, coding help, careful explanations and office-style work.
Total72 / 100
Relevance
Why this matters for beginners.
USA - led by Dario Amodei
Claude is strong for writing, long documents, coding help, careful explanations and office-style work. The score below is the sum of ten visible categories. Each category is worth 10 points.
LLMDocumentsCodingComputer use
No fudging
Ten category scores.
The total score is calculated by adding these ten rows. Change a total only by changing one or more category scores and writing down why.
| Category | Score | Why it sits here | What would change it |
|---|---|---|---|
| Beginner usefulness | 8 / 10 | Strong everyday utility for non-technical users. | Move this if the product becomes easier or harder for a normal beginner to use. |
| Cost clarity | 7 / 10 | Costs can be understood, but plan limits need checking. | Move this when pricing, credit rules, refunds or cancellation paths change. |
| Data and privacy | 8 / 10 | Public privacy controls and business settings are comparatively clear. | Move this when privacy, retention, training or human-review terms change. |
| Terms and rights | 7 / 10 | Terms are usable but not simple enough to skim. | Move this when output rights, commercial use or upload licences change. |
| Transparency | 8 / 10 | Official docs, model pages or policy material are comparatively visible. | Move this when official docs, model cards, incident notes or pricing pages improve or disappear. |
| Safety record | 8 / 10 | Public safety work or abuse handling is comparatively strong. | Move this when safety policies, misuse incidents, scams or consent controls materially change. |
| Leadership ethics | 7 / 10 | Leadership and incentive context is mixed. | Move this when leadership, governance, controversies or investment conflicts change. |
| Defence involvement | 3 / 10 | Anthropic has direct defence exposure through a CDAO award and says Claude is deployed across U.S. defence and national-security work. It scores slightly above OpenAI here because Anthropic publicly refused to remove safeguards on mass domestic surveillance and fully autonomous weapons, but the defence involvement itself is still substantial.Sources: CDAO frontier AI partnerships | Anthropic Department of War statement | Anthropic Acceptable Use Policy | Move this when Anthropic publishes clearer deployment boundaries, contract scope, surveillance safeguards or autonomous-weapons safeguards. |
| Country and stability | 7 / 10 | Country and data-jurisdiction context should be considered before sensitive use. | Move this when data jurisdiction, surveillance context, censorship or political stability changes. |
| Public benefit | 9 / 10 | There is visible public-benefit signal such as open releases, science, education or accessibility. | Move this when public-interest work, open releases, science or accessibility evidence changes. |
Plain signals
What was counted.
These are the visible card notes from the main comparison page, kept here so the quick card and the working page stay connected.
Trust watch
Direct CDAO award, defence and national-security deployments, partner delivery through Palantir/AWS, copyright litigation and policy complexity.
Good signs
Public refusal to remove mass domestic-surveillance and fully autonomous-weapons safeguards, plus strong safety research and enterprise privacy options.
Research links
Sources to refresh first.
Start with official product, pricing, privacy and policy pages. Then check public reporting for major controversies, ownership, defence work and legal disputes.