Tested 7 d.Revised Apr. 26
Tool review · Detailed verdict
Anthropic
Claude Opus 4.7
« First choice for writing and code. With Claude Design (April 2026), Anthropic also moves onto Figma's turf. »
Overall score
weighted average · 5 axes
§ Presentation video · Claude Opus 4.7
YouTube · officiel
§ Product preview
Price / month
20 €
Pro plan
Cost / 1M tokens
15 €
output rate
Context
500k
max tokens
GDPR
8.4/10
compliant
Hosting
🇺🇸 US
by default
§3.1Strength profile
Scale 0 to 10 · RadarOnAI internal method
§3.2+ Strong points
- 01Most natural writing style
- 02Claude Design: prototypes/slides from a prompt
- 03Better-calibrated refusals
§3.3− Weak points
- 01No image output
- 02Variable queue time at peak hours
Best for
Consultants, writers, developers who need nuance.
Not for
Anyone after massive multimodal or visual creation.
§04Benchmarks — Claude Opus 4.7
Against the rest, cold.
BenchmarkScoreValueRank
GPQA Diamond
Scientific reasoning
82.4%
#1
SWE-bench Verified
Real-world bug fixing
74.1%
#1
MMLU-Pro
General knowledge
83.9%
#3
HumanEval+
Code generation
94.2%
#1
FR-Check 2026
Internal language test
91/100
#1
Overall score evolution · Q1 '25 → Q3 '26
Q1 '25Q2 '25Q3 '25Q4 '25Q1 '26Q2 '26Q3 '26
+29
pts in 7 quarters
§ Editor's view · try it?
If your long-form text matters more than your images, Claude Opus 4.7 remains the subscription that pays back its 20 € monthly fee fastest. We use it daily in the newsroom.
Affiliate link ↗ — RadarOnAI earns a commission if you subscribe, at no extra cost to you. Our verdicts never depend on commissions.
Essayer Claude ↗
No commitment · cancel in one click