The Open-Source AI Stack
RSS
All models

Models · Compare

Llama 4 Maverick vs Gemini 2.5 Pro

Rows highlighted in warm gray are where the models differ. Numbers carry their as-of date and primary source.

Specs

Field A: Llama 4 Maverick B: Gemini 2.5 Pro
Released 2025-04-052025-03-25
Developer MetaGoogle DeepMind
Openness Source-availableProprietary
License Llama 4 Community LicenseProprietary
OSI-approved nono
Data released nono
Training code nono
Architecture moeunknown
Total params 400B
Active params 17B
Experts 128 (1 active)
Context window 1.0M
Attention gqaunknown
Position enc. rope-llama3unknown
Pretraining tokens
Post-training sft, online-rl, dporlhf
Training hardware H100
$/M input $0.35$1.25
$/M output $0.85$10.00
Output tok/sec 108.9127.3

Benchmarks

Missing scores render as not reported; never inferred. Bold highlights the leader per benchmark.

General reasoning

MMLU-Pro 86.2 2026-05-21

Code

LiveCodeBench 80.1 2026-05-21

Math

MATH 88.9 2026-05-21 96.7 2026-05-21
AIME 2024 39.0 2026-05-21
AIME 2025 19.3 2026-05-21 87.7 2026-05-21

Held-out / arena

LMArena Elo 1417.0 2025-04-05

Context · A

The mid-tier sibling in Meta's first MoE Llama family, sharing the natively multimodal early-fusion architecture used across Llama 4. Maverick pairs 17B active parameters with 128 routed experts, and Meta positioned it against GPT-4o and Gemini 2.0 Flash on standard reasoning and coding suites.

Context · B

The first Gemini release to clearly lead on LMArena Elo and on hard reasoning benchmarks. Native 1M-token context with reported 2M expansion in pipeline.

Llama 4 Maverick detail → · Gemini 2.5 Pro detail →