SOURCE-REVIEWED MODEL PROFILE
Mistral-Small-3.1-24B-Instruct-2503
A 24B multimodal instruction model supporting text and image understanding, with a 128K context window.
Metadata reviewed 2026-10-24 · Original creator: mistralai
What it does
- Text and image input
- Instruction following and function calling
- Long-context conversations
Local fit
Large GPU / more RAM. Try brainstorming from an image. Image input needs the correct vision files and a supporting backend.
Configuration value; backend support and enough memory are required. Start with a smaller context on limited hardware.
Evidence and limits
Source reviewed; no local generation benchmark for this exact checkpoint.
Connect a supported local text/vision backend; not independently tested
License and source
Keep upstream notices and check the exact downloaded artifact and parent-model terms. The original creator remains credited.
Read the original model card ↗
Reviewed source revision: 68faf511d618ef198fef186659617cfd2eb8e33a