Claude review & score
A strong coder with a 200k context and sharp writing. But performance is unstable run-to-run, and in our critical testing it scored very poorly on security-sensitive tasks. No native image generation, and a smaller feature spread.
$20 / mo · top tier $100–200 (Max) · updated 14 June 2026 · prices approx.
Claude by category
How Claude scores in each weighted category (0–100). Yellow bar = category winner overall.
Where Claude shines — and where it doesn't
- General reasoning
- Coding
- Long-context handling
- Instruction following / steerability
- Creative quality
- Prompt-quality tolerance
- Image generation
- Image editing
- Video understanding
- Multiple providers in one interface
- Side-by-side answer comparison
Claude technical specs & limits
Approximate; verify on the official page.
| Flagship model | Latest Claude |
|---|---|
| Context window | 200K tokens |
| Max output (approx.) | ~32–64K tokens |
| Standard usage limit (approx.) | ~45 msgs / 5h |
| Max file upload (approx.) | ~30 MB |
| Files per chat (approx.) | ~20 |
| Supported file types | PDF, text, CSV, images |
| Image generation limit | — |
| Live web access | limited |
| Languages | 95+ |
Claude plans
$0
Free tier to try it out, with limits.
$20 / mo
The main paid plan most people choose.
$100–200 (Max)
For power users and teams.
Official source: Anthropic pricing · last checked 14 June 2026. Prices and limits change often — confirm before subscribing.
Is Claude worth it?
A strong coder with a 200k context and sharp writing. But performance is unstable run-to-run, and in our critical testing it scored very poorly on security-sensitive tasks. No native image generation, and a smaller feature spread.
Claude – common questions
How much does Claude cost?
Claude starts at $20 / mo for its main paid plan (Pro); the top tier is $100–200 (Max). A free tier is available. Prices are approximate – check the official page.
What is Claude best at?
In our scoring Claude is strongest where it wins or rates highly; see its category scorecard above. Overall it scores 80.2/100 (rank #4).
Is Claude worth it?
A strong coder with a 200k context and sharp writing. But performance is unstable run-to-run, and in our critical testing it scored very poorly on security-sensitive tasks. No native image generation, and a smaller feature spread.