Benchfolio

Leaderboard / Model

GPT-5.6 Terra

OpenAI · Closed weights · Released Jul 9, 2026 · $2 in · $12 out per 1M tokens

52Overall · rank 16

Score by job

Coding

65 · rank 14 · 3 of 5 benchmarks

BenchmarkResultVariantz
DeepSWE69.6%gpt-5.6-terra_max0.80
FrontierCode41.3%gpt-5.6-terra_unknown0.18
LMArena Coding1485 ratinggpt-5.6-terra-xhigh-0.26

Agents

Not enough data · 1 of 4 benchmarks

BenchmarkResultVariantz
APEX-Agents58.2%gpt-5.6-terra_max0.56

Reasoning

49 · rank 19 · 3 of 4 benchmarks

BenchmarkResultVariantz
GPQA Diamond93.3%gpt-5.6-terra_max0.70
ARC-AGI-283.9%gpt-5.6-terra_max0.92
SimpleBench48.9%gpt-5.6-terra_unknown-1.65

Math

68 · rank 8 · 5 of 5 benchmarks

BenchmarkResultVariantz
FrontierMath Tiers 1–386.0%gpt-5.6-terra_max1.24
FrontierMath Tier 470.7%gpt-5.6-terra_max1.24
OTIS Mock AIME99.7%gpt-5.6-terra_max1.02
ProofBench74.0%gpt-5.6-terra_xhigh1.04
LMArena Math1460 ratinggpt-5.6-terra-xhigh-0.77

Writing

46 · rank 36 · 2 of 2 benchmarks

BenchmarkResultVariantz
LMArena Creative Writing1409 ratinggpt-5.6-terra-xhigh-1.26
LMArena Instruction Following1453 ratinggpt-5.6-terra-xhigh-0.43

Design

33 · rank 28 · 1 of 1 benchmarks

BenchmarkResultVariantz
LMArena WebDev1519 ratinggpt-5.6-terra-xhigh (codex-harness)-0.39