GPT-5.4 Pro vs Llama 4 Scout 17B 16e Instruct
As of Oct 15, 2026, GPT-5.4 Pro leads Llama 4 Scout 17B 16e Instruct in 2 of 2 shared categories: Science and Abstraction (ARC).
GPT-5.4 Pro
It used to be. Was SOTA in Math.
Llama 4 Scout 17B 16e Instruct
#90 Best recorded rank: #90 in Abstraction (ARC), as of Oct 15, 2026.
| Category | GPT-5.4 Pro | Llama 4 Scout 17B 16e Instruct |
|---|---|---|
| Reasoning | #— · Not enough data Margin — Epoch AI benchmarking hub · Mar 5, 2026 | #— · Not enough data Margin — Epoch AI benchmarking hub · Apr 5, 2025 |
| Math | #6 · Was SOTA · Math Margin 11.111111111111114 Epoch AI benchmarking hub · Mar 5, 2026 | — |
| Science | #6 · #6 in Science Margin — Epoch AI benchmarking hub · Mar 5, 2026 | #122 · #122 in Science Margin — Epoch AI benchmarking hub · Apr 5, 2025 |
| Vision | — | #— · Not enough data Margin — LMArena leaderboard dataset · Oct 9, 2026 |
| Factuality | #— · Not enough data Margin — Epoch AI benchmarking hub · Mar 5, 2026 | — |
| Abstraction (ARC) | #8 · #8 in Abstraction (ARC) Margin — Epoch AI benchmarking hub · Mar 5, 2026 | #90 · #90 in Abstraction (ARC) Margin — Epoch AI benchmarking hub · Apr 5, 2025 |
GPT-5.4 Pro: $67.5 per 1M blended tokens.
Llama 4 Scout 17B 16e Instruct: price per 1M tokens is unset in the catalog.
GPT-5.4 Pro: closed weights.
Llama 4 Scout 17B 16e Instruct: open weights.
Science: As of Oct 15, 2026, GPT-5.4 Pro leads Llama 4 Scout 17B 16e Instruct in Science at #6 against #122.
Abstraction (ARC): As of Oct 15, 2026, GPT-5.4 Pro leads Llama 4 Scout 17B 16e Instruct in Abstraction (ARC) at #8 against #90.
GPT-5.4 Pro · Llama 4 Scout 17B 16e Instruct
GPT-5.4 Pro or Llama 4 Scout 17B 16e Instruct?
As of Oct 15, 2026, GPT-5.4 Pro leads Llama 4 Scout 17B 16e Instruct in 2 of 2 shared categories: Science and Abstraction (ARC).
GPT-5.4 Pro or Llama 4 Scout 17B 16e Instruct for Science?
As of Oct 15, 2026, GPT-5.4 Pro leads Llama 4 Scout 17B 16e Instruct in Science at #6 against #122.
GPT-5.4 Pro or Llama 4 Scout 17B 16e Instruct for Abstraction (ARC)?
As of Oct 15, 2026, GPT-5.4 Pro leads Llama 4 Scout 17B 16e Instruct in Abstraction (ARC) at #8 against #90.