Quick start

ollama run minicpm-v:latest

Runnable variants

Memory is a rough file-size estimate; context length, runtime and GPU offload change it.

12 of 17 matching variants shown.17 of 17 variants shown.

Tag	Type	Inputs	Digest	Download	Parameters	Quantization	Context	Est. memory	Run
8b	Typelocal	Inputstext, image	Digestc92bfad01205	Download5.5 GB	Parameters8b	QuantizationUnknown	Context32K	Est. memory≈ 7.5 GB	`ollama run minicpm-v:8b`
8b-2.6-fp16	Typelocal	Inputstext, image	Digestf3f122c78635	Download16 GB	Parameters8b	QuantizationFP16	Context32K	Est. memory≈ 19 GB	`ollama run minicpm-v:8b-2.6-fp16`
8b-2.6-q2_K	Typelocal	Inputstext, image	Digest61cf7cf5edda	Download4.1 GB	Parameters8b	QuantizationQ2_K	Context32K	Est. memory≈ 6 GB	`ollama run minicpm-v:8b-2.6-q2_K`
8b-2.6-q3_K_L	Typelocal	Inputstext, image	Digesta62e8056b1f0	Download5.1 GB	Parameters8b	QuantizationQ3_K_L	Context32K	Est. memory≈ 7 GB	`ollama run minicpm-v:8b-2.6-q3_K_L`
8b-2.6-q3_K_M	Typelocal	Inputstext, image	Digestb7c14f8aad8c	Download4.9 GB	Parameters8b	QuantizationQ3_K_M	Context32K	Est. memory≈ 6.5 GB	`ollama run minicpm-v:8b-2.6-q3_K_M`
8b-2.6-q3_K_S	Typelocal	Inputstext, image	Digest0e5dd84b9d34	Download4.5 GB	Parameters8b	QuantizationQ3_K_S	Context32K	Est. memory≈ 6.5 GB	`ollama run minicpm-v:8b-2.6-q3_K_S`
8b-2.6-q4_0	Typelocal	Inputstext, image	Digestc92bfad01205	Download5.5 GB	Parameters8b	QuantizationQ4_0	Context32K	Est. memory≈ 7.5 GB	`ollama run minicpm-v:8b-2.6-q4_0`
8b-2.6-q4_1	Typelocal	Inputstext, image	Digest1e17cbffa666	Download5.9 GB	Parameters8b	QuantizationQ4_1	Context32K	Est. memory≈ 8 GB	`ollama run minicpm-v:8b-2.6-q4_1`
8b-2.6-q4_K_M	Typelocal	Inputstext, image	Digest950a671abea6	Download5.7 GB	Parameters8b	QuantizationQ4_K_M	Context32K	Est. memory≈ 7.5 GB	`ollama run minicpm-v:8b-2.6-q4_K_M`
8b-2.6-q4_K_S	Typelocal	Inputstext, image	Digesteda89b106de0	Download5.5 GB	Parameters8b	QuantizationQ4_K_S	Context32K	Est. memory≈ 7.5 GB	`ollama run minicpm-v:8b-2.6-q4_K_S`
8b-2.6-q5_0	Typelocal	Inputstext, image	Digest4b28c460b85e	Download6.4 GB	Parameters8b	QuantizationQ5_0	Context32K	Est. memory≈ 8.5 GB	`ollama run minicpm-v:8b-2.6-q5_0`
8b-2.6-q5_1	Typelocal	Inputstext, image	Digest6ac2a6038e89	Download6.8 GB	Parameters8b	QuantizationQ5_1	Context32K	Est. memory≈ 9 GB	`ollama run minicpm-v:8b-2.6-q5_1`
8b-2.6-q5_K_M	Typelocal	Inputstext, image	Digest6dfa70b0f281	Download6.5 GB	Parameters8b	QuantizationQ5_K_M	Context32K	Est. memory≈ 8.5 GB	`ollama run minicpm-v:8b-2.6-q5_K_M`
8b-2.6-q5_K_S	Typelocal	Inputstext, image	Digestb4d3d0c08b9c	Download6.4 GB	Parameters8b	QuantizationQ5_K_S	Context32K	Est. memory≈ 8.5 GB	`ollama run minicpm-v:8b-2.6-q5_K_S`
8b-2.6-q6_K	Typelocal	Inputstext, image	Digestac20b8a12390	Download7.3 GB	Parameters8b	QuantizationQ6_K	Context32K	Est. memory≈ 9.5 GB	`ollama run minicpm-v:8b-2.6-q6_K`
8b-2.6-q8_0	Typelocal	Inputstext, image	Digest9e2efdd7b657	Download9.1 GB	Parameters8b	QuantizationQ8_0	Context32K	Est. memory≈ 11.5 GB	`ollama run minicpm-v:8b-2.6-q8_0`
latestdefault	Typelocal	Inputstext, image	Digestc92bfad01205	Download5.5 GB	ParametersUnknown	QuantizationUnknown	Context32K	Est. memory≈ 7.5 GB	`ollama run minicpm-v:latest`

Related models

bakllavalocal

BakLLaVA is a multimodal model consisting of the Mistral 7B base model augmented with the LLaVA architecture.

d1local

d1 is a 3B decision model from Liquid AI, built on LFM2.5-VL-3B

translategemmalocal

A new collection of open translation models built on Gemma 3, helping people communicate across 55 languages.

medgemmalocal

MedGemma is a collection of Gemma 3 variants that are trained for performance on medical text and image comprehension.