1
gemini-3.1-pro-previewokSevere
google
88.0
Model Arena · depth shallow
Middle East & North Africa · 2026-07-09
One day, one region, several language models — each handed exactly the same headlines and readings, then asked how tense things would be a week later. Because the input is identical, any difference between their answers is a difference between the models.
This page shows what they said. Whether any of it was right is a separate question, answered on the Leaderboard — and the short version there is that none of them beat guessing.
How 'identical evidence' is guaranteed
Every model in a run receives the same rendered prompt over the same evidence payload, and the payload's hash is stored with the run and displayed below. If two models had been shown different material the hashes would differ, so the fairness of the comparison is checkable rather than asserted.
Models that crashed, timed out, or returned unparseable output are persisted as rows and ranked here alongside the successes.
Consensus
70.0
trimmed median over 5 ok score(s); 2 trimmed
Completion
100.0%
5 scored / 5 attempted — failures stay in the denominator
Evidence
07bda1a4…d2a6
10 headline(s), shared by every model
Ranking
prompt p1depth shallow
Models that failed are listed here too, marked and unscored, rather than quietly dropped. A model that only answers when the question is easy would otherwise look like a model that is good at the question — every leaderboard that hides its failures is really reporting "accuracy among the runs that happened to work".
2
deepseek-reasonerokSevere
deepseek
78.0
3
claude-opus-5okSevere
anthropic
70.0
4
llama-3.3-70b-versatileokUnstable
groq
70.0
5
claude-haiku-4-5okUnstable
anthropic
58.0
Every model in this run returned a parseable verdict, so no failure row is shown. The failure states are rendered by the same component and are covered by tests/model_row_test.tsx.