Contamination MapText / LLM / Model
Llama 4 (Scout, Maverick)
- Organisation
- Meta
- Family
- Llama · version 4
- Released
- 2025-04-05 (Model card 'Model Release Date: April 5, 2025'; Meta blog dated April 5, 2025)
- Training cutoff
- 2024-08-01
Evidence
“The pretraining data has a cutoff of August 2024”
raw.githubusercontent.com · model card · Training Data / Data Freshness · retrieved 2026-10-15 - Checkpoints
- meta-llama/Llama-4-Scout-17B-16E, meta-llama/Llama-4-Scout-17B-16E-Instruct, meta-llama/Llama-4-Maverick-17B-128E, meta-llama/Llama-4-Maverick-17B-128E-Instruct, meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8
- Weights
- Open · Llama 4 Community License Agreement
- Links
- Model cardCode
Training data disclosure: Vague
Level determined for 4% of benchmark cells.
- TestTest data seen0
- SeenSeen, split unclear0
- TrainTraining data seen0
- Related data seen0
- NoNot seen1
- ?Unknown21
“A mix of publicly available, licensed data and information from Meta’s products and services. This includes publicly shared posts from Instagram and Facebook and people’s interactions with Meta AI.”
“The overall data mixture for training consisted of more than 30 trillion tokens, which is more than double the Llama 3 pre-training mixture and includes diverse text, image, and video datasets.”
Other declared training data
-
publicly available data (broad web data: may contain benchmark test sets) · split: unknown
Scout ~40T tokens, Maverick ~22T tokens of multimodal data (model card).
Evidence
“A mix of publicly available, licensed data and information from Meta’s products and services. This includes publicly shared posts from Instagram and Facebook and people’s interactions with Meta AI.”
raw.githubusercontent.com · model card · Model Information table · retrieved 2026-10-15 -
licensed data (not yet matched to a dataset) · split: unknown
Licensors not named.
Evidence
“A mix of publicly available, licensed data and information from Meta’s products and services. This includes publicly shared posts from Instagram and Facebook and people’s interactions with Meta AI.”
raw.githubusercontent.com · model card · Model Information table · retrieved 2026-10-15 -
information from Meta's products and services (public Instagram/Facebook posts, interactions with Meta AI) (broad web data: may contain benchmark test sets) · split: unknown
User-generated public posts can contain arbitrary text.
Evidence
“A mix of publicly available, licensed data and information from Meta’s products and services. This includes publicly shared posts from Instagram and Facebook and people’s interactions with Meta AI.”
raw.githubusercontent.com · model card · Model Information table · retrieved 2026-10-15 -
human-generated data from vendors combined with synthetic data (not yet matched to a dataset) · split: unknown · fine-tuning
Evidence
“We employ a multi-faceted approach to data collection, combining human-generated data from our vendors with synthetic data to mitigate potential safety risks.”
raw.githubusercontent.com · model card · Fine-tuning data · retrieved 2026-10-15
Benchmark cells
Classic LLM benchmarks
Unknown MMLU (Measuring Massive Multitask Language Understanding) · training data not disclosed in enough detail
Unknown GSM8K (Grade School Math 8K) · training data not disclosed in enough detail
Unknown MATH (Mathematics Aptitude Test of Heuristics) · training data not disclosed in enough detail
Unknown HumanEval · training data not disclosed in enough detail
Unknown ARC-Challenge (AI2 Reasoning Challenge, Challenge Set) · training data not disclosed in enough detail
Unknown HellaSwag · training data not disclosed in enough detail
Unknown TruthfulQA · training data not disclosed in enough detail
Unknown WinoGrande · training data not disclosed in enough detail
Multimodal LLM benchmarks
Unknown MMMU (Massive Multi-discipline Multimodal Understanding and Reasoning) · training data not disclosed in enough detail
Unknown MMMU-Pro · training data not disclosed in enough detail
Unknown MathVista · training data not disclosed in enough detail
Unknown ChartQA · training data not disclosed in enough detail
Unknown DocVQA · training data not disclosed in enough detail
Unknown AI2D (AI2 Diagrams) · training data not disclosed in enough detail
Unknown CharXiv · training data not disclosed in enough detail
Recent LLM benchmarks
Unknown GPQA Diamond · training data not disclosed in enough detail
Unknown MMLU-Pro · training data not disclosed in enough detail
Unknown LiveCodeBench (code generation, release_v6) · training data not disclosed in enough detail
Unknown SWE-bench Verified · training data not disclosed in enough detail
Unknown AIME 2024 · training data not disclosed in enough detail
Not seen AIME 2025 · its test items were published from 2025-02-06, after the model's stated training-data cutoff (2024-08-01)
Evidence (3)
“The pretraining data has a cutoff of August 2024”
“2025 AIME I problems and solutions. The test was held on Thursday, February 6, 2025.”
“This year the AIME I will take place on February 6, 2025, while the alternate AIME II will take place on February 12, 2025.”
Unknown MATH-500 · training data not disclosed in enough detail
Notes
Scout and Maverick share the description and cutoff but differ in token counts (~40T vs ~22T). The blog describes Maverick as codistilled from the unreleased Llama 4 Behemoth. No Llama release newer than Llama 4 found in the meta-llama Hugging Face organisation or the llama-models repository as of 2026-10-15. Accepts image input: “The Llama 4 collection of models are natively multimodal AI models that enable text and multimodal experiences.” Accepts image input: “Llama 4 has been tested for image understanding up to 5 input images.”
Sources
- Llama 4 MODEL_CARD.md (meta-llama/llama-models) · model card · retrieved 2026-10-15
- The Llama 4 herd (Meta AI blog, April 5, 2025) · website · retrieved 2026-10-15
- meta-llama/llama-models README (launch dates) · readme · retrieved 2026-10-15