See how a question turns into research.

Read published Briefs to see the question, signals, and output. Use the benchmark to compare how models perform on the same task.

Benchmark

How we test model output.

The benchmark runs the same research task across models so you can compare coverage, agreement, and output quality.

Open benchmark
Benchmark · July 11, 2026
Top model
Claude Opus-4.7
Anthropic · 86.0 composite
Models tested
34
Industries
12
Free scan

Bring your next question into Signals.

Start with the theme behind the next decision.