Mixtral 8x7B

Mistral AI · Mixtral family · released 2023-12-11 · Apache-2.0 license

Mixtral 8x7B by Mistral AI — 46.7B MoE (12.9B active) open-weight local LLM. License: Apache-2.0. Benchmarks, VRAM requirements per quantization, GPU compatibility, pricing and community reviews on slopsome.com.

Key specs

TypeLocal open-weight
Parameters46.7B total · MoE, 12.9B active
Architecture
Context window33K tokens
Knowledge cutoff
Modalitiestext
Recommended backends
Minimum viable rig

Benchmark scores

GPQA Diamond38%
SWE-bench Verified18%
AIME28%
MMLU-Pro56%
BFCL v3 (tool use)64%
Composite score13.03
Community ratingNo reviews yet

VRAM & disk per quantization

QuantVRAMDiskRAMContext
Q4_K_M28 GB27 GB32 GB33K