Claude Opus 4 vs Claude Opus 5
As of Oct 14, 2026, Claude Opus 4 leads Claude Opus 5 in 0 of 5 shared categories.
Claude Opus 4
It used to be. Was SOTA in Reasoning.
Claude Opus 5
Close. Near-SOTA in Factuality, rank #4.
| Category | Claude Opus 4 | Claude Opus 5 |
|---|---|---|
| General LLM | #102 · #102 in General LLM Margin 69.38775510204081 LMArena leaderboard dataset · Oct 13, 2026 | #11 · Was SOTA · General LLM Margin 7.482993197278915 LMArena leaderboard dataset · Oct 13, 2026 |
| Reasoning | #97 · Was SOTA · Reasoning Margin 67.12328767123287 LMArena leaderboard dataset · Oct 13, 2026 | #9 · Was SOTA · Reasoning Margin 6.849315068493155 LMArena leaderboard dataset · Oct 13, 2026 |
| Math | #29 · #29 in Math Margin 62.22222222222222 Epoch AI benchmarking hub · May 22, 2025 | #— · Not enough data Margin — LMArena leaderboard dataset · Oct 13, 2026 |
| Science | #91 · #91 in Science Margin — Epoch AI benchmarking hub · May 22, 2025 | #12 · #12 in Science Margin — Epoch AI benchmarking hub · Jul 24, 2026 |
| Coding | #— · Not enough data Margin — Epoch AI benchmarking hub · May 22, 2025 | #4 · #4 in Coding Margin 22.727272727272734 Terminal-Bench Harbor Hub · Oct 6, 2026 |
| Coding › Agentic | — | #4 · #4 in Coding › Agentic Margin 22.727272727272734 Terminal-Bench Harbor Hub · Oct 6, 2026 |
| Coding › Web dev | — | #6 · Was SOTA · Coding › Web dev Margin 4.672897196261687 LMArena leaderboard dataset · Oct 13, 2026 |
| Long context | #70 · #70 in Long context Margin 47.61904761904762 LMArena leaderboard dataset · Oct 13, 2026 | #9 · Was SOTA · Long context Margin 6.122448979591837 LMArena leaderboard dataset · Oct 13, 2026 |
| Vision | #— · Not enough data Margin — LMArena leaderboard dataset · Oct 9, 2026 | #12 · #12 in Vision Margin 36.8421052631579 Epoch AI benchmarking hub · Jul 24, 2026 |
| Factuality | — | #4 · Near-SOTA · Factuality Margin 3.731343283582092 LMArena leaderboard dataset · Oct 13, 2026 |
| Abstraction (ARC) | #69 · #69 in Abstraction (ARC) Margin — Epoch AI benchmarking hub · May 22, 2025 | #3 · #3 in Abstraction (ARC) Margin — Epoch AI benchmarking hub · Jul 24, 2026 |
Claude Opus 4: price per 1M tokens is unset in the catalog.
Claude Opus 5: $10 per 1M blended tokens.
Claude Opus 4: closed weights.
Claude Opus 5: closed weights.
General LLM: As of Oct 14, 2026, Claude Opus 5 leads Claude Opus 4 in General LLM at #11 against #102.
Reasoning: As of Oct 14, 2026, Claude Opus 5 leads Claude Opus 4 in Reasoning at #9 against #97.
Science: As of Oct 14, 2026, Claude Opus 5 leads Claude Opus 4 in Science at #12 against #91.
Long context: As of Oct 14, 2026, Claude Opus 5 leads Claude Opus 4 in Long context at #9 against #70.
Abstraction (ARC): As of Oct 14, 2026, Claude Opus 5 leads Claude Opus 4 in Abstraction (ARC) at #3 against #69.
Claude Opus 4 or Claude Opus 5?
As of Oct 14, 2026, Claude Opus 4 leads Claude Opus 5 in 0 of 5 shared categories.
Claude Opus 4 or Claude Opus 5 for General LLM?
As of Oct 14, 2026, Claude Opus 5 leads Claude Opus 4 in General LLM at #11 against #102.
Claude Opus 4 or Claude Opus 5 for Reasoning?
As of Oct 14, 2026, Claude Opus 5 leads Claude Opus 4 in Reasoning at #9 against #97.
Claude Opus 4 or Claude Opus 5 for Science?
As of Oct 14, 2026, Claude Opus 5 leads Claude Opus 4 in Science at #12 against #91.