ai.canshu.com
中文
VRAM Calculator Advisor & TCO API Pricing Compare Models Leaderboard GPUs

Search models, GPUs, and tools

↑↓ Navigate · Enter Open · ★ Watchlist prioritized ai.canshu.com
Alibaba Qwen
Open-Weight Alibaba Qwen · qwen2-5-omni-7b · Text+Image+Audio+Video

Qwen2.5-Omni 7B Official Specifications, Pricing & Benchmarks

Qwen2.5-Omni 7B (qwen2-5-omni-7b) by Alibaba Qwen, released 2024-12, featuring 32,768 context window, 2,048 max output tokens, and Text+Image+Audio+Video input.

📋 Qwen2.5-Omni 7B Official API & System Card Specifications

Verified against official Alibaba Qwen developer documentation & pricing

API Model ID: qwen2-5-omni-7b

1. Identity & Capabilities

API Spec
API Model ID
qwen2-5-omni-7b
Release Date
2024-12
Weight License
Open-Weight (API Spec)
Native Reasoning
—
Function Calling & JSON Schema
✅ 原生支持
Batch API 异步批处理
✅ Supported (50% off)

2. Context, Output & Modalities

Limits
Max Context Window
33,000 tokens (33K)
Max Output Tokens
2,048 tokens
Input Modalities
Text + Image + Audio + Video
Output Modalities
Text + Audio
Direct China Access
✅ Direct CN + RMB
Deployment Mode
Open-Weight + Cloud API

3. Official Pricing & Cache

USD / 1M
Input / 1M Tokens
$ 0.1
Cached Input / 1M
$ 0.02 -80%
Output / 1M Tokens
$ 0.4

🎛️ Official Sampling & Reasoning Parameters

Recommended generation settings & API parameter constraints

Standard Generation Config
temperature=0.7 top_p=0.95 max_tokens=2,048

📊 Qwen2.5-Omni 7B Full Benchmark Scorecard

30 public benchmarks · ranks among 154 models · synced 2026-11-20

⚔️ Compare head-to-head

📐 Math 7 items

GSM8k #3/4 88.7%

Grade School Math 8K, a dataset of 8.5K high-quality linguistically diverse grade school math word problems requiring multi-step reasoning and elementary arithmetic operations.

MATH #6/9 71.5%

MATH dataset contains 12,500 challenging competition mathematics problems from AMC 10, AMC 12, AIME, and other mathematics competitions.

MMLU-Redux #15/15 71%

An improved version of the MMLU benchmark featuring manually re-annotated questions to identify and correct errors in the original dataset. Provides more reliable evaluation metrics for language models by addressing dataset quality issues found in the original MMLU.

MathVista #7/7 67.9%

MathVista evaluates mathematical reasoning of foundation models in visual contexts. It consists of 6,141 examples derived from 28 existing multimodal datasets and 3 newly created datasets (IQTest, FunctionQA, and PaperQA), combining challenges from diverse mathematical and…

MMLU-Pro #39/39 47%

A more robust and challenging multi-task language understanding benchmark that extends MMLU by expanding multiple-choice options from 4 to 10, eliminating trivial questions, and focusing on reasoning-intensive tasks.

LiveBench #26/27 29.6%

LiveBench is a challenging, contamination-limited LLM benchmark that addresses test set contamination by releasing new questions monthly based on recently-released datasets, arXiv papers, news articles, and IMDb movie synopses.

MathVision #20/20 25%

MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities.

👁️ Multimodal 17 items

DocVQA #1/3 95.2%

A dataset for Visual Question Answering on document images containing 50,000 questions defined on 12,000+ document images.

ChartQA #3/3 85.3%

ChartQA is a large-scale benchmark comprising 9.6K human-written questions and 23.1K questions generated from human-written chart summaries, designed to evaluate models' abilities in visual and logical reasoning over charts.

TextVQA #1/1 84.4%

TextVQA contains 45,336 questions on 28,408 images that require reasoning about text to answer. Introduced to benchmark VQA models' ability to read and reason about text within images, particularly for assistive technologies for visually impaired users.

AI2D #7/7 83.2%

AI2D is a dataset of 4,903 illustrative diagrams from grade school natural sciences (such as food webs, human physiology, and life cycles) with over 15,000 multiple choice questions and answers.

MMBench-V1.1 #6/6 81.8%

Version 1.1 of MMBench, an improved bilingual benchmark for assessing multi-modal capabilities of vision-language models through multiple-choice questions in both English and Chinese, providing systematic evaluation across diverse vision-language tasks.

VideoMME w sub. #8/8 72.4%

The first-ever comprehensive evaluation benchmark of Multi-modal LLMs in Video analysis. Features 900 videos (254 hours) with 2,700 question-answer pairs covering 6 primary visual domains and 30 subfields.

MVBench #7/7 70.3%

A comprehensive multi-modal video understanding benchmark covering 20 challenging video tasks that require temporal understanding beyond single-frame analysis.

RealWorldQA #13/13 70.3%

RealWorldQA is a benchmark designed to evaluate basic real-world spatial understanding capabilities of multimodal models.

EgoSchema #1/1 68.6%

A diagnostic benchmark for very long-form video language understanding consisting of over 5000 human curated multiple choice questions based on 3-minute video clips from Ego4D, covering a broad range of natural human activities and behaviors.

MMAU #4/4 65.6%

A massive multi-task audio understanding and reasoning benchmark comprising 10,000 carefully curated audio clips paired with human-annotated natural language questions spanning speech, environmental sounds, and music.

MMStar #8/8 64%

MMStar is an elite vision-indispensable multimodal benchmark comprising 1,500 challenge samples meticulously selected by humans to evaluate 6 core capabilities and 18 detailed axes.

MuirBench #3/3 59.2%

A comprehensive benchmark for robust multi-image understanding capabilities of multimodal LLMs. Consists of 12 diverse multi-image tasks involving 10 categories of multi-image relations (e.g., multiview, temporal relations, narrative, complementary).

MMMU #16/16 59.2%

MMMU (Massive Multi-discipline Multimodal Understanding) is a benchmark designed to evaluate multimodal models on college-level subject knowledge and deliberate reasoning.

OCRBench_V2 #4/5 57.8%

OCRBench v2: Enhanced large-scale bilingual benchmark for evaluating Large Multimodal Models on visual text localization and reasoning with 10,000 human-verified question-answering pairs across 8 core OCR capabilities.

ODinW #6/7 42.4%

Object Detection in the Wild (ODinW) benchmark for evaluating object detection models' task-level transfer ability across diverse real-world datasets in terms of prediction accuracy and adaptation efficiency.

MMMU-Pro #42/44 36.6%

A more robust multi-discipline multimodal understanding benchmark that enhances MMMU through a three-step process: filtering text-only answerable questions, augmenting candidate options, and introducing vision-only input settings.

MM-MT-Bench #1/1 0.1

A multi-turn LLM-as-a-judge evaluation benchmark for testing multimodal instruction-tuned models' ability to follow user instructions in multi-turn dialogues and answer open-ended questions in a zero-shot manner.

Source: ZeroEval / llm-stats.com (✓ = third-party verified, others vendor-reported).

Closest Self-Hostable Open-Weight Alternatives to Qwen2.5-Omni 7B

Open-weight models with full HuggingFace weights for private GPU / Mac deployment

Compare with these →

Frequently Asked Questions

Is Qwen2.5-Omni 7B open-source or proprietary?

Qwen2.5-Omni 7B (qwen2-5-omni-7b) is a proprietary commercial model by Alibaba Qwen. Weights and internal layer dimensions are undisclosed; it is accessed via cloud API.

What are the context limit and API pricing of Qwen2.5-Omni 7B?

Qwen2.5-Omni 7B supports 33K input context and 2,048 max output tokens, priced at $0.1/1M input ($0.02/1M cached) and $0.4/1M output.