Seed 2.0 Mini Official Specifications, Pricing & Benchmarks
Seed 2.0 Mini (doubao-seed-2-0-mini-260428) by ByteDance 字节豆包, released 2026-02-14, featuring 256,000 context window, 131,072 max output tokens, and Text+Image+Video input.
📋 Seed 2.0 Mini Official API & System Card Specifications
Verified against official ByteDance 字节豆包 developer documentation & pricing
1. Identity & Capabilities
API Spec- API Model ID
- doubao-seed-2-0-mini-260428
- Release Date
- 2026-02-14
- Weight License
- Proprietary (Undisclosed)
- Native Reasoning
- ✅ Supported
- 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
- 256,000 tokens (256K)
- Max Output Tokens
- 131,072 tokens
- Input Modalities
- Text + Image + Video
- Output Modalities
- Text
- Direct China Access
- ✅ Direct CN + RMB
- Deployment Mode
- Cloud API only (Proprietary)
3. Official Pricing & Cache
CNY / 1M- Input / 1M Tokens
- ¥ 0.2
- Cached Input / 1M
- ¥ 0.04 -80%
- Output / 1M Tokens
- ¥ 2
🎛️ Official Sampling & Reasoning Parameters
Recommended generation settings & API parameter constraints
📊 Seed 2.0 Mini Full Benchmark Scorecard
80 public benchmarks · ranks among 154 models · synced 2026-11-20
💻 Coding 2 items
AetherCode is a competitive-programming benchmark of olympiad-level algorithmic coding problems.
CL-bench is an open-source benchmark with its own data and rubrics for evaluating models on coding and agentic tasks, scored using a setup fully aligned with the official procedure.
📐 Math 14 items
All 30 problems from the 2025 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999.
All 30 problems from the 2026 American Invitational Mathematics Examination (AIME I and AIME II), testing olympiad-level mathematical reasoning with integer answers from 000-999.
MathVista-Mini is a smaller version of the MathVista benchmark that evaluates mathematical reasoning in visual contexts.
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.
Multilingual Massive Multitask Language Understanding dataset released by OpenAI, featuring professionally translated MMLU test questions across 14 languages including Arabic, Bengali, German, Spanish, French, Hindi, Indonesian, Italian, Japanese, Korean, Portuguese, Swahili…
MATH-Vision is a dataset designed to measure multimodal mathematical reasoning capabilities.
Beyond AIME is a difficult mathematical reasoning benchmark designed to test deeper reasoning chains and harder decomposition than standard AIME-style problem sets.
SuperGPQA is a comprehensive benchmark that evaluates large language models across 285 graduate-level academic disciplines.
A multimodal mathematics and reasoning benchmark focused on dynamic visual problem solving.
EMMA (Enhanced MultiModal reAsoning) is a benchmark for organic multimodal reasoning across mathematics, physics, chemistry, and coding.
A competitive programming benchmark using problems from the CodeForces platform. The benchmark evaluates code generation capabilities of LLMs on algorithmic problems with difficulty ratings ranging from 800 to 2400.
FrontierScience Olympiad is a benchmark of olympiad-level scientific reasoning problems, testing expert understanding and multi-step reasoning across advanced natural-science domains.
Text-only Humanity's Last Exam variant evaluated without tool use.
MathArena Apex is a challenging math contest benchmark featuring the most difficult mathematical problems designed to test advanced reasoning and problem-solving abilities of AI models. It focuses on olympiad-level mathematics and complex multi-step mathematical reasoning.
👁️ Multimodal 46 items
A vision-language benchmark that probes blind spots and brittle reasoning in multimodal models.
CharXiv-D is the descriptive questions subset of the CharXiv benchmark, designed to assess multimodal large language models' ability to extract basic information from scientific charts.
TempCompass is a comprehensive benchmark for evaluating temporal perception capabilities of Video Large Language Models (Video LLMs).
RealWorldQA is a benchmark designed to evaluate basic real-world spatial understanding capabilities of multimodal models.
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.
Video-MMMU evaluates Large Multimodal Models' ability to acquire knowledge from expert-level professional videos across six disciplines through three cognitive stages: perception, comprehension, and adaptation.
MMMU (Massive Multi-discipline Multimodal Understanding) is a benchmark designed to evaluate multimodal models on college-level subject knowledge and deliberate reasoning.
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.
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).
LongVideoBench is a question-answering benchmark featuring video-language interleaved inputs up to an hour long.
BLINK: Multimodal Large Language Models Can See but Not Perceive. A benchmark for multimodal language models focusing on core visual perception abilities.
LiveSports-3K evaluates fine-grained understanding and commentary of live sports video.
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.
CharXiv-R is the reasoning component of the CharXiv benchmark, focusing on complex reasoning questions that require synthesizing information across visual chart elements.
TVBench is a temporal video understanding benchmark evaluating reasoning over actions, events, and temporal dynamics in videos.
OVOBench (Online Video Online Benchmark) evaluates streaming video understanding, testing a model's ability to perceive and respond to video content in real time as it unfolds.
MMVU (Multimodal Multi-disciplinary Video Understanding) is a benchmark for evaluating multimodal models on video understanding tasks across multiple disciplines, testing comprehension and reasoning capabilities on video content.
DUDE (Document Understanding Dataset and Evaluation) tests multi-page, multi-domain document understanding and reasoning.
VideoSimpleQA evaluates factual knowledge grounded in video content, measuring how accurately models answer short, fact-seeking questions about videos.
LVBench is an extreme long video understanding benchmark designed to evaluate multimodal models on videos up to two hours in duration. It contains 6 major categories and 21 subcategories, with videos averaging five times longer than existing datasets.
ChartQAPro is a challenging benchmark for question answering over diverse, real-world charts and infographics.
A comprehensive benchmark designed to evaluate image-context reasoning in large visual-language models (LVLMs) by challenging models with 346 images and 1,129 carefully crafted questions to assess language hallucination and visual illusion.
MotionBench is a benchmark for evaluating multimodal models on motion understanding in videos, testing the ability to comprehend temporal dynamics, movement patterns, and action sequences.
OVBench is an online video understanding benchmark that evaluates a model's ability to perceive, memorize, and reason about real-time video streams as they unfold.
VideoHolmes evaluates video understanding and reasoning capabilities in multimodal models.
CrossVid evaluates cross-video reasoning, requiring models to integrate information across multiple videos.
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.
VLMsAreBiased evaluates whether vision-language models rely on visual evidence or fall back on language priors when answering.
TreeBench evaluates visual grounded reasoning, requiring models to localize and reason about fine-grained visual details.
Embodied Reasoning Question Answering benchmark consisting of 400 multiple-choice visual questions across spatial reasoning, trajectory reasoning, action reasoning, state estimation, and multi-view reasoning for evaluating AI capabilities in physical world interactions.
ContPhy is a continuum physical-reasoning benchmark evaluating understanding of physical dynamics in video.
Minerva is a benchmark for complex video reasoning, evaluating models on multi-step reasoning over long and information-dense video content.
WorldVQA is a benchmark designed to evaluate atomic vision-centric world knowledge. It assesses models' ability to understand and reason about visual elements representing real-world knowledge.
TOMATO (Temporal Reasoning Multimodal Evaluation) assesses multimodal models on motion and temporal perception in video, testing understanding of actions, motion, and changes over time.
A benchmark for early-stage visual reasoning and perception on child-like vision tasks.
The Abstraction and Reasoning Corpus for Artificial General Intelligence (ARC-AGI) is a benchmark designed to test general intelligence and abstract reasoning capabilities through visual grid-based transformation tasks.
VisFactor is a benchmark evaluating fine-grained visual factor perception and reasoning over images.
OmniDocBench 1.5 is a comprehensive benchmark for evaluating multimodal large language models on document understanding tasks, including OCR, document parsing, information extraction, and visual question answering across diverse document types.
ARC-AGI-2 is an upgraded benchmark for measuring abstract reasoning and problem-solving abilities in AI systems through visual grid transformation tasks.
CountBench evaluates object counting capabilities in visual understanding.
SimpleVQA is a visual question answering benchmark focused on simple queries.
RefSpatialBench evaluates spatial reference understanding and grounding.
MMLongBench-Doc evaluates long document understanding capabilities in vision-language models.
VisuLogic evaluates logical reasoning capabilities in visual contexts.
ZEROBench-Sub is a subset of the ZEROBench benchmark.
ZEROBench is a challenging vision benchmark designed to test models on zero-shot visual understanding tasks.
📜 Long Context 2 items
MRCR v2 (8-needle) is a variant of the Multi-Round Coreference Resolution benchmark that includes 8 needle items to retrieve from long contexts. This tests models' ability to simultaneously track and reason about multiple pieces of information across extended conversations.
MRCR v2 (Multi-Round Coreference Resolution version 2) is an enhanced version of the synthetic long-context reasoning task.
🧠 Reasoning & Science 13 items
A graph reasoning benchmark that evaluates language models' ability to find parent nodes in graphs with context length under 128k tokens, requiring understanding of graph structure and edge relationships.
COLLIE is a grammar-based framework for systematic construction of constrained text generation tasks.
Global PIQA is a multilingual commonsense reasoning benchmark that evaluates physical interaction knowledge across 100 languages and cultures.
A challenging dataset of 448 multiple-choice questions written by domain experts in biology, physics, and chemistry. Questions are Google-proof and extremely difficult, with PhD experts reaching 65% accuracy.
A graph reasoning benchmark that evaluates language models' ability to perform breadth-first search (BFS) operations on graphs with context length under 128k tokens, returning nodes reachable at specified depths.
LiveCodeBench is a holistic and contamination-free evaluation benchmark for large language models for code.
MultiChallenge is a realistic multi-turn conversation evaluation benchmark that challenges frontier LLMs across four key categories: instruction retention (maintaining instructions throughout conversations), inference memory (recalling and connecting details from previous…
PHYBench is a benchmark of real-world physics problems spanning mechanics, electromagnetism, thermodynamics, optics, and modern physics, designed to evaluate physical perception and multi-step quantitative reasoning in large language models.
A fine-grained atomic evaluation metric for factual precision in long-form text generation that breaks generated text into atomic facts and computes the percentage supported by reliable knowledge sources, with automated assessment using retrieval and language models.
SimpleQA Verified is a curated, reliability-focused subset of SimpleQA that addresses label noise and redundancy in the original benchmark, measuring short-form parametric factual accuracy of large language models on fact-seeking questions with single, indisputable answers.
SuperChem is a benchmark of advanced chemistry problems requiring expert-level domain knowledge and reasoning.
FrontierScience Research is a benchmark evaluating AI models on cutting-edge scientific research questions requiring deep domain expertise, multi-step reasoning, and synthesis of complex scientific concepts across disciplines.
ARC-AGI-2 is the second-generation Abstraction and Reasoning Corpus benchmark measuring fluid, general reasoning and abstraction.
🏥 Domains 2 items
An open-source benchmark for measuring performance and safety of large language models in healthcare, consisting of 5,000 multi-turn conversations evaluated by 262 physicians using 48,562 unique rubric criteria across health contexts and behavioral dimensions.
A challenging variation of HealthBench that evaluates large language models' performance and safety in healthcare through 5,000 multi-turn conversations with particularly rigorous evaluation criteria validated by 262 physicians from 60 countries.
Source: ZeroEval / llm-stats.com (✓ = third-party verified, others vendor-reported).
Closest Self-Hostable Open-Weight Alternatives to Seed 2.0 Mini
Open-weight models with full HuggingFace weights for private GPU / Mac deployment
Frequently Asked Questions
Is Seed 2.0 Mini open-source or proprietary?
Seed 2.0 Mini (doubao-seed-2-0-mini-260428) is a proprietary commercial model by ByteDance 字节豆包. Weights and internal layer dimensions are undisclosed; it is accessed via cloud API.
What are the context limit and API pricing of Seed 2.0 Mini?
Seed 2.0 Mini supports 256K input context and 131,072 max output tokens, priced at ¥0.2/1M input (¥0.04/1M cached) and ¥2/1M output.