GPT-5.6 Terra Pro
OpenAI / Jul 9, 2026 / openrouter
1 judged output · judge spread σ 0.52 · W 0 / L 3 / D 0
Leaderboard
One output per model, independent judge scores, and bounded close-score pairwise weighting. The best Gary Busey portraits rise to the top.
Official standings
OpenAI / Jul 9, 2026 / openrouter
1 judged output · judge spread σ 0.52 · W 0 / L 3 / D 0
OpenAI / Jul 9, 2026 / openrouter
1 judged output · judge spread σ 0.88 · W 8 / L 0 / D 1
Google / Feb 25, 2026 / openrouter
1 judged output · judge spread σ 0.85 · W 5 / L 0 / D 4
Anthropic / Jul 24, 2026 / openrouter
1 judged output · judge spread σ 1.25 · W 4 / L 3 / D 2
Anthropic / Jul 24, 2026 / openrouter
1 judged output · judge spread σ 0.89 · W 7 / L 1 / D 4
Google / Feb 19, 2026 / openrouter
1 judged output · judge spread σ 0.86 · W 3 / L 3 / D 4
OpenAI / Apr 24, 2026 / openrouter
1 judged output · judge spread σ 0.62 · W 6 / L 3 / D 3
OpenAI / Jul 9, 2026 / openrouter
1 judged output · judge spread σ 0.87 · W 2 / L 6 / D 4
MoonshotAI / Jul 16, 2026 / openrouter
1 judged output · judge spread σ 0.66 · W 0 / L 9 / D 0
Google / May 19, 2026 / openrouter
1 judged output · judge spread σ 1.16 · W 1 / L 2 / D 6
Anthropic / Jun 9, 2026 / openrouter
1 judged output · judge spread σ 0.79 · W 3 / L 5 / D 1
OpenAI / Jul 9, 2026 / openrouter
1 judged output · judge spread σ 1.04 · W 2 / L 3 / D 3
Meta / Aug 5, 2026 / openrouter
1 judged output · judge spread σ 0.74 · W 4 / L 2 / D 2
Meta / Unknown / direct api
1 judged output · judge spread σ 0.83 · W 2 / L 2 / D 4
Anthropic / Feb 4, 2026 / openrouter
1 judged output · judge spread σ 1.37 · W 4 / L 1 / D 3
OpenAI / Dec 10, 2025 / openrouter
1 judged output · judge spread σ 0.47 · W 2 / L 4 / D 3
Qwen / Aug 12, 2026 / openrouter
1 judged output · judge spread σ 1.35 · W 6 / L 1 / D 6
MoonshotAI / Apr 20, 2026 / openrouter
1 judged output · judge spread σ 0.73 · W 8 / L 1 / D 3
Anthropic / Nov 24, 2025 / openrouter
1 judged output · judge spread σ 0.80 · W 8 / L 2 / D 4
Anthropic / May 27, 2026 / openrouter
1 judged output · judge spread σ 0.40 · W 2 / L 7 / D 1