GPT-5.4 Pro vs Llama 4 Maverick 17B 128e Instruct
As of Oct 14, 2026, GPT-5.4 Pro leads Llama 4 Maverick 17B 128e Instruct in 2 of 2 shared categories: Math and Abstraction (ARC).
GPT-5.4 Pro
It used to be. Was SOTA in Math.
Llama 4 Maverick 17B 128e Instruct
It used to be. Was SOTA in Math.
| Category | GPT-5.4 Pro | Llama 4 Maverick 17B 128e 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 | #37 · #37 in Math Margin 82.22222222222223 Epoch AI benchmarking hub · Apr 5, 2025 |
| Science | #6 · #6 in Science Margin — Epoch AI benchmarking hub · Mar 5, 2026 | — |
| 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 | #88 · #88 in Abstraction (ARC) Margin — Epoch AI benchmarking hub · Apr 5, 2025 |
GPT-5.4 Pro: $67.5 per 1M blended tokens.
Llama 4 Maverick 17B 128e Instruct: price per 1M tokens is unset in the catalog.
GPT-5.4 Pro: closed weights.
Llama 4 Maverick 17B 128e Instruct: open weights.
Math: As of Oct 14, 2026, GPT-5.4 Pro leads Llama 4 Maverick 17B 128e Instruct in Math at #6 against #37.
Abstraction (ARC): As of Oct 14, 2026, GPT-5.4 Pro leads Llama 4 Maverick 17B 128e Instruct in Abstraction (ARC) at #8 against #88.
GPT-5.4 Pro · Llama 4 Maverick 17B 128e Instruct
GPT-5.4 Pro or Llama 4 Maverick 17B 128e Instruct?
As of Oct 14, 2026, GPT-5.4 Pro leads Llama 4 Maverick 17B 128e Instruct in 2 of 2 shared categories: Math and Abstraction (ARC).
GPT-5.4 Pro or Llama 4 Maverick 17B 128e Instruct for Math?
As of Oct 14, 2026, GPT-5.4 Pro leads Llama 4 Maverick 17B 128e Instruct in Math at #6 against #37.
GPT-5.4 Pro or Llama 4 Maverick 17B 128e Instruct for Abstraction (ARC)?
As of Oct 14, 2026, GPT-5.4 Pro leads Llama 4 Maverick 17B 128e Instruct in Abstraction (ARC) at #8 against #88.