Llama 4 Maverick

Meta · released 2025-04-05 · Llama 4 Community license

Llama 4 Maverick by Meta — 400B MoE (17B active) open-weight local LLM. License: Llama 4 Community. Benchmarks, VRAM requirements per quantization, GPU compatibility, pricing and community reviews on slopsome.com.

Key specs

TypeLocal open-weight
Parameters400B total · MoE, 17B active
ArchitectureMoE (128 experts)
Context window
Knowledge cutoff2024-08-01
Modalitiestext, image
Recommended backends
Minimum viable rig

Benchmark scores

GPQA Diamond69%
SWE-bench Verified43%
AIME60%
MMLU-Pro81%
BFCL v3 (tool use)68%
Composite score6.06
Community ratingNo reviews yet

VRAM & disk per quantization

QuantVRAMDiskRAMContext
Q4_K_M205 GB200 GB256 GB1000K
Q8405 GB400 GB448 GB1000K

API pricing (per 1M tokens)

ProviderInputOutputFree tier
Groq$0.2$0.6No
Fireworks AI$0.22$0.88No
Together AI$0.27$0.85No