# Scores transcribed from the model card coding and multimodal benchmark charts.
name = "MiMo-V2.5"
description = "Open MiMo model for multimodal coding agents and long-context automation"
family = "mimo"
release_date = "2026-04-22"
last_updated = "2026-04-22"
attachment = true
reasoning = true
temperature = true
tool_call = true
knowledge = "2024-12"
open_weights = true

[limit]
context = 1_048_576
output = 131_072

[modalities]
input = ["text", "image", "audio", "video"]
output = ["text"]

[[weights]]
label = "Hugging Face"
url = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "MiMo Coding Bench"
score = 71.8
metric = "score"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "Claw-Eval"
score = 62.3
metric = "score"
variant = "text"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "Terminal-Bench"
score = 65.8
metric = "score"
version = "2.0"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "SWE-Bench Pro"
score = 56.1
metric = "score"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "CharXiv Reasoning"
score = 81
metric = "score"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "MMMU-Pro"
score = 77.9
metric = "score"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "HR-Bench"
score = 88.5
metric = "score"
variant = "4K"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "OmniDocBench"
score = 87.2
metric = "score"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "Claw-Eval"
score = 23.8
metric = "score"
variant = "multimodal"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "Video-MME"
score = 87.7
metric = "score"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "DailyOmni"
score = 83.5
metric = "score"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"

[[benchmarks]]
name = "VideoHolmes"
score = 64
metric = "score"
source = "https://huggingface.co/XiaomiMiMo/MiMo-V2.5"