← All verdicts

Which AI can you trust for catalog lookup?

By , Founder and Self-Directed Space Researcher, Viventine Space Systems · updated

Look up a specific object: NORAD and COSPAR IDs, owners, launch and decay dates, orbit class.

On their own, no AI earns "Trust it" here. With Orbit Sentinel tools: Claude Opus 5.5, Claude Fable 5.1, Claude Sonnet 5.5, Claude Haiku 5.5, Grok 4.7.

AI Closed-book With Orbit Sentinel tools n
right / wrong range declined right / wrong range declined
Claude Opus 5.5 Anthropic
67 / 11
52–79 22
100 / 0
92–100 0 45×3
Claude Fable 5.1 Anthropic
64 / 7
50–77 29
99 / 1
90–100 0 45×2
Claude Sonnet 5.5 Anthropic
58 / 24
43–71 18
98 / 2
88–100 0 45×3
Claude Haiku 5.5 Anthropic
29 / 18
18–43 53
91 / 1
79–96 7 45×3
Grok 4.7 xAI
47 / 0
33–61 53
91 / 1
79–96 8 45×2
Gemma 4 12B Google
0 / 68
0–9 32
78 / 15
64–87 8 45×3
gpt-oss 20B OpenAI
7 / 62
2–18 31
76 / 10
61–86 14 45×3
DeepSeek-R1 14B DeepSeek
2 / 76
0–12 22
—
45×1

% right / % wrong per cell; color is the verdict (green trust it, amber verify, red don't use). Range is the 95% interval for % right, counting each question once; n is questions × runs for the tools track.

Sample questions

All published questions with their answers →