v2.2 · 2026-06-12
VISION / MULTIMODAL MODEL — RUNTIME REQUIREMENTS
·RAM and VRAM estimates are conservative local-inference estimates based on GGUF file size plus runtime and KV-cache overhead, not official Mistral requirements.
·Vision support requires a runtime that supports the model's multimodal inputs; text-only runtimes may expose only text chat.
Mistral Small 3.2 24B Instruct 2506
mistral-small-3-2-24b-instruct-2506
Mistral Small 3.2 24B Instruct 2506 is a 24B text chat model profile for local AI planning. This page records provider, license, context, quantization, hardware-fit estimates, setup hints, source links, and caveats so readers do not choose a model by name alone.
PROVIDER
Mistral AI
FAMILY
mistral
MODEL TYPE
text-chat
PARAMETERS
24B
MODALITIES
text, image
ARCHITECTURE
decoder-only Transformer in the Mistral Small 3.2 instruct family with image-text input support
CONTEXT WINDOW
128,000 tokens
TRAINING TOKENS
—
RELEASE DATE
2025-06-20
LICENSE & LINKS
Apache-2.0license text
GGUF REPOSITORIES
bartowski/mistralai_Mistral-Small-3.2-24B-Instruct-2506-GGUF community conversion
Community GGUF conversion with Q4_K_M, Q5_K_M and Q8_0 file-size table; notes template handling for tool calling.
unsloth/Mistral-Small-3.2-24B-Instruct-2506-GGUF community conversion
Community GGUF repo documenting chat-template fixes and tool-calling behavior.
SETUP HINTS
OLLAMA
ollama run mistral-small3.2:24b-instruct-2506-q4_K_MLM STUDIO
Use bartowski or lmstudio-community GGUF builds for Mistral-Small-3.2-24B-Instruct-2506; select Q4_K_M on 24 GB VRAM or 32 GB unified-memory systems.
LLAMA.CPP
Use llama.cpp with the Q4_K_M GGUF and the Mistral Small 3.2 Jinja chat template when tool calling is required.QUANTIZATION FILE SIZES (GGUF)
Q4_K_M
14.33 GB
Q5_K_M
16.76 GB
Q8_0
25.05 GB
Quantization note: Q4_K_M
RAM / VRAM ESTIMATES
MIN RAM
20 GB
COMFORTABLE RAM
32 GB
MIN VRAM
18 GB
COMFORTABLE VRAM
24 GB
These are conservative local-inference estimates, not official hardware requirements.
HARDWARE FIT
CPU only (no GPU)
limited
8 GB RAM
no
16 GB RAM
limited
32 GB RAM
comfortable
8 GB VRAM
no
12 GB VRAM
partial_offload_only
24 GB VRAM
good
Apple Silicon (unified memory)
comfortable_on_32gb_or_more_unified_memory
Hardware fit values are conservative local-inference estimates based on GGUF size plus runtime and KV-cache overhead. Actual requirements depend on context length, quantization, runtime, GPU offload, and other running apps.
BEST FOR
·high-quality local 24B-class chat
·vision-capable local assistant workflows
·function-calling and instruction-following use
·single-4090 or 32 GB unified-memory class deployments
AVOID IF
·you need a text-only model card with no multimodal runtime caveats
·you have 16 GB RAM or less
·your runtime cannot apply the required chat template for tool calling
CAVEATS
·RAM and VRAM estimates are conservative local-inference estimates based on GGUF file size plus runtime and KV-cache overhead, not official Mistral requirements.
·This target uses Mistral Small 3.2 rather than 3.1 because the 3.2 2506 artifact has a clear official model card, current Ollama tag, and multiple GGUF listings.
·Vision support requires a runtime that supports the model's multimodal inputs; text-only runtimes may expose only text chat.
·Some GGUF notes indicate tool calling depends on using the correct Jinja chat template.
·Official Mistral documentation lists this 2506 model with a deprecation date of 2026-04-30 and a replacement path to Mistral Small 4 for hosted/API model selection; local GGUF artifacts remain available.
SOURCE URLS
FIELD EVIDENCE
contextWindowTokensdocs.mistral.ai/models/model-cards/mistral-small-3-2-25-06
LANGUAGES
EnglishFrenchGermanSpanishItalianPortugueseJapaneseKoreanRussianChineseArabicPersianIndonesianMalayNepaliPolishRomanianSerbianSwedishTurkishUkrainianVietnameseHindiBengali
CAPABILITIES
tool callingvision / image input