MMODELYST
Models/Llama 4 Scout
MDL

Llama 4 Scout

Meta

Verified against Artificial Analysis · Jul 13, 2026
Capability
43
percentile index
Price /M
$0.17 / $0.66
in / out
Context
10M
tokens
Avg score
33.3
11 benchmarks
AA Intelligence
10.0
index
Speed
107 tok/s
TTFT 0.6s
Released
Apr 5, 2025
About

Llama 4 Scout is a open-weights model in the Llama family from Meta. Benchmarked on 11 evals, averaging 33.3.

Benchmark scores · 11
avg 33.3 — higher is better
MATH
84.4
MMLU-Pro
75.2
GPQA Diamond
58.7
IFBench
39.5
LiveCodeBench
29.9
AA Long-Context Reasoning
25.8
SciCode
17
τ²-bench
15.5
AIME
14
Humanity's Last Exam
4.3
Terminal-Bench Hard
1.5
Details
PublisherMeta
FamilyLlama
Parameters
License
WeightsOpen
ReleasedApr 5, 2025
Performance & price
Output speed107 tok/s
Latency (TTFT)0.6s
Price in / out$0.17 / $0.66 per 1M
Blended price$0.292 / 1M
Coding index8.2
Math index14.0

Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →

Capability over time

Append-only ledger — every observed change to this model's numbers.

API providers
Artificial Analysis$0.66/M