Llama 4 Maverick vs Llama 4 Scout

Benchmarks, API pricing and specs, head to head. Data updated 2026-06-10.

Llama 4 Maverick

Meta · Apr 2025

Meta's natively multimodal 400B MoE — the largest openly downloadable US-made model, served cheaply across many hosts.

Llama 4 Scout

Meta · Apr 2025

A 10M-token context window — still the largest ever shipped — in a 109B MoE that fits on a single H100.

The verdict

Llama 4 Maverick wins 3 of the 3 benchmarks these models share, against 0 for Llama 4 Scout. Llama 4 Scout is about 1.5x cheaper per blended million tokens (3:1 input:output mix). Llama 4 Scout also takes 10M of context versus 1M for Llama 4 Maverick.

Benchmark head-to-head 30

GPQA Diamond
69.8%57.2%
Llama 4 MaverickLlama 4 Scout
MMMU
73.4%69.4%
Llama 4 MaverickLlama 4 Scout
MMLU-Pro
80.5%74.3%
Llama 4 MaverickLlama 4 Scout

Specs & pricing

Llama 4 MaverickLlama 4 Scout
modhub Index
Input price / 1M$0.27$0.18
Output price / 1M$0.85$0.59
Context window1M10M
Max output
Open weightsyes (Llama 4 Community License)yes (Llama 4 Community License)
Reasoning modelnono
Multimodal inputtext, imagetext, image
Knowledge cutoffAug 2024Aug 2024
ReleasedApr 2025Apr 2025
Example monthly cost*$3.98$2.68

* 10M input + 1.5M output tokens per month at list prices, no caching. Green = better value on that row.

Frequently asked questions

Which is better, Llama 4 Maverick or Llama 4 Scout?
Llama 4 Maverick wins 3 of the 3 benchmarks these models share, against 0 for Llama 4 Scout. Llama 4 Scout is about 1.5x cheaper per blended million tokens (3:1 input:output mix). Llama 4 Scout also takes 10M of context versus 1M for Llama 4 Maverick.
Which is cheaper, Llama 4 Maverick or Llama 4 Scout?
Llama 4 Maverick costs $0.27/$0.85 per million input/output tokens, while Llama 4 Scout costs $0.18/$0.59. For a typical workload of 10M input and 1.5M output tokens per month, that's $3.98 versus $2.68.
Which model is better for coding, Llama 4 Maverick or Llama 4 Scout?
We don't yet track SWE-bench Verified results for both models; check their individual pages for coding-related scores.

More comparisons