THE COMPARISON WORKSPACE
Understand the tradeoffs.
Choose two or three models. Start with the evidence, then tune your priorities.
Loading comparison data…
Your next model starts with a comparison.
Add candidates from the catalog, or choose models above.
Find models →Decision brief
Preference score uses the selected measured dimensions, normalized by their total weight. It is a heuristic, not a success probability. Missing dimensions prevent a score.
Coverage counts measured dimensions, not confidence. Benchmark conditions and public source links may be missing. Listed dataset prices; provider terms and dates need verification. Zero without explicit free-price evidence is treated as unknown.
Key differences
This is index variability, not API uptime or reliability.
Advanced preferences
Preference score uses the selected measured dimensions, normalized by their total weight. It is a heuristic, not a success probability. Missing dimensions prevent a score.