VENICE ATLAS · MODEL DOSSIER · MIMO-V2.5

Model overview

MiMo-V2.5

Name: MiMo-V2.5
Brand: venice.ai
Price: 0.4000 USD
Availability: InStock

MiMo-V2.5 is Xiaomi's native omnimodal model with strong agentic capabilities, supporting text, image, video, and audio understanding in a unified architecture. Built on a sparse Mixture-of-Experts backbone with 310B total and 15B active parameters, it delivers long-context reasoning up to 1M tokens, function calling, and multimodal perception.

xiaomi-mimo-v2-5 · venice.ai

Text · Venice catalog$0.40 per 1M tokensPrice on Venice AI

ModalityTextOn Venice AI

Primary price$0.40 per 1M tokensper 1M tokens

Context window1MMax tokens

Benchmarks14Benchmark results

Benchmark results

See the tests behind each score. They use the benchmark’s original scoring scale.

Video understanding#108.31 rating

Agents2 results

Finance Agent v2finance-agent-v2 · agents

36.7%

ReportedOct 2026

ResearchClawBenchresearchclawbench · agents

16.9%

ReportedOct 2026

Coding2 results

Claw-Evalclaw-eval · code

63.2%

ReportedOct 2026

MiMo Coding Benchmimo-coding-bench · code

71.8%

ReportedOct 2026

Long Context1 result

GraphWalksgraphwalks · long_context

87.0%

ReportedOct 2026

Multimodal1 result

Video-MMEvideo-mme · multimodal

87.7%

ReportedOct 2026

Reasoning3 results

CharXiv-Rcharxiv-r · reasoning

81.0%

ReportedOct 2026

SWE-Bench Proswe-bench-pro · reasoning

56.1%

ReportedOct 2026

Terminal-Bench 2.0terminal-bench-2 · reasoning

65.8%

ReportedOct 2026

Video understanding2 results

DailyOmnidailyomni · video

83.5%

ReportedOct 2026

VideoHolmesvideoholmes · video

64.0%

ReportedOct 2026

Vision3 results

HR-Bench (4k)hr-bench-4k · vision

88.5%

ReportedOct 2026

MMMU-Prommmu-pro · vision

77.9%

ReportedOct 2026

OmniDocBenchomnidocbench · vision

87.2%

ReportedOct 2026

Learn what each benchmark measures ↗

Pricing detail

Input price per 1M tokens: $0.40
Output price per 1M tokens: $2.00
Cache read price per 1M tokens: $0.0800
Blended price per 1M tokens (50/50): $1.20

Prices are published by Venice AI. Venice model page ↗

Capabilities

ReasoningFunction callingVisionAudio inputVideo inputWeb search

Endpoint metadata

Privacy: private
Quantization: fp8
Maximum completion tokens: 66K
Uncensored: No
Beta: No
Mapping confidence: 100.0%

View model source ↗

All capabilities

Max Images: 10
Quantization: fp8
Supports E2 E E: No
Supports Vision: Yes
Supports X Search: No
Optimized For Code: Yes
Supports Log Probs: No
Supports Reasoning: Yes
Supports Web Search: Yes
Supports Audio Input: Yes
Supports Video Input: Yes
Default Reasoning Effort: high
Reasoning Effort Options: none, low, medium, high
Supports Multiple Images: Yes
Supports Response Schema: Yes
Supports Tee Attestation: No
Supports Function Calling: Yes
Supports Reasoning Effort: Yes

Related models

More from xiaomi

xiaomi-mimo-v2-6-flashMiMo-V2.6-Flash$0.17 input per 1M tokens ↗