Seed 2.1 Turbo
ByteDanceProprietary
विवरण
ByteDance's low-cost, low-latency next-generation agent model for large-scale production. A deep-thinking model with the full feature set of Seed 2.1 Pro and comparable performance, it is designed for enterprise-grade deployments handling high volumes of online calls across coding, long-chain agents, and multimodal tasks. Served via Volcano Engine as Doubao-Seed-2.1-turbo.
रिलीज़ तिथि
2026-06-24
पैरामीटर
—
संदर्भ लंबाई
262K
मोडैलिटीज़
image, text, video
क्षमता रडार
60
general
50
coding
60
reasoning
51
scienceअनुमानित
70
agents
70
multimodal
समर्पित विज्ञान बेंचमार्क उपलब्ध न होने पर Science का अनुमान LLM Stats विज्ञान स्कोर या तर्क से लगाया जाता है।
रैंकिंग
| डोमेन | #रैंक | स्कोर | स्रोत |
|---|---|---|---|
| एजेंटिक क्षमता | 61 | 50.0 | LS |
| मल्टीमॉडल रैंकिंग | 25 | 58.0 | LS |
बेंचमार्क स्कोर (LLM Stats)
(LLM Stats (zeroeval))3d
BLINK
79.4%स्वयं
Agents
BrowseCompOpenAI (2025)
84.9%स्वयं
GDPval
82.7%स्वयं
MCP Atlas
80.3%स्वयं
OSWorld
76.4%स्वयं
Web Bench
73.6%स्वयं
OfficeQA Pro
71.1%स्वयं
MobileWorld
70.0%स्वयं
Terminal-Bench 2.1
67.6%स्वयं
CyberGym
67.0%स्वयं
OneMillion Bench
66.6%स्वयं
SeedClawBench
63.8%स्वयं
WildClawBench
62.8%स्वयं
Trae Code Gen
59.7%स्वयं
SWE-Bench ProPrinceton NLP (2024)
57.0%स्वयं
Trae Error Fix
56.7%स्वयं
xDailyBench
56.4%स्वयं
Finance Agent v1.1
56.0%स्वयं
Workspace Bench
54.7%स्वयं
Agent Startup Bench
54.0%स्वयं
Program Bench
49.4%स्वयं
Toolathlon
49.1%स्वयं
Doubao Multi-Turn Bench
49.0%स्वयं
PresentBench
48.3%स्वयं
Repo Env
46.7%स्वयं
ClawEval-MM
46.0%स्वयं
NL2Repo
43.7%स्वयं
CreativeWork
34.5%स्वयं
SWE-Atlas
30.6%स्वयं
APEX-Agents
29.2%स्वयं
GameWorld
25.9%स्वयं
DeepSWE
23.0%स्वयं
PostTrainBench
18.3%स्वयं
Biology
SciCode
57.8%स्वयं
Chemistry
SuperGPQA
67.4%स्वयं
SuperChem
56.6%स्वयं
Code
FrontierCS
50.8%स्वयं
Artifacts Bench
47.0%स्वयं
Image2FloorPlan
35.9%स्वयं
Coding
AetherCode
67.9%स्वयं
Embodied
EmbSpatialBench
0.82 / 100स्वयं
General
MMMU-Pro
82.2%स्वयं
SimpleVQA
0.71 / 100स्वयं
KINA
46.6%स्वयं
MSQA
42.0%स्वयं
Image To Text
OCRBench_V2
62.8%स्वयं
Knowledge
VideoSimpleQA
71.4%स्वयं
WorldBench
63.7%स्वयं
Long Context
DUDE
83.1%स्वयं
LongVideoBench
80.6%स्वयं
MMLongBench-128K
76.9%स्वयं
LVBench
76.8%स्वयं
Math
MathVision
92.7%स्वयं
MathVista
90.5%स्वयं
MathVerse
89.2%स्वयं
Beyond AIME
88.0%स्वयं
EMMA
78.4%स्वयं
FrontierScience Olympiad
76.0%स्वयं
DynaMath
68.1%स्वयं
Humanity's Last Exam
54.6%स्वयं
MathArena Apex
35.4%स्वयं
LiveMathematicianBench
27.7%स्वयं
HorizonMath
2.0%स्वयं
Multimodal
CharXiv-D
94.6%स्वयं
Video-MME
89.0%स्वयं
CharXiv-R
83.6%स्वयं
OVOBench
79.2%स्वयं
TVBench
77.2%स्वयं
LiveSports-3K
77.1%स्वयं
MotionBench
74.8%स्वयं
TreeBench
71.1%स्वयं
ChartQAPro
70.9%स्वयं
OVBench
69.7%स्वयं
VLMsAreBiased
68.3%स्वयं
VideoHolmes
67.6%स्वयं
Minerva
65.9%स्वयं
CrossVid
63.2%स्वयं
BabyVision
62.9%स्वयं
ContPhy
61.1%स्वयं
MeasureBench
58.9%स्वयं
TOMATO
56.8%स्वयं
VisuLogic
0.53 / 100स्वयं
WorldVQA
48.6%स्वयं
VisFactor
43.9%स्वयं
MMSIBench
31.4%स्वयं
ZEROBench
0.11 / 100स्वयं
Reasoning
ERQA
71.3%स्वयं
ArcAGI2
61.3%स्वयं
FrontierScience Research
33.3%स्वयं
Spatial Reasoning
RealWorldQA
86.3%स्वयं
AA मूल्यांकन सूचकांक
(Artificial Analysis)कोई AA मूल्यांकन डेटा उपलब्ध नहीं
LLM Stats श्रेणी स्कोर
(LLM Stats (zeroeval))Structured Output90
Legal80
Search80
Spatial Reasoning80
Embodied80
Finance80
3d80
Long Context70
Multimodal70
Safety70
Image To Text70
Healthcare70
Economics70
Tool Calling70
Video70
Vision70
Math60
Physics60
Reasoning60
Knowledge60
General60
Agents60
Biology60
Chemistry60
Coding60
Science50
Frontend Development50
Code50
Systems20
मूल्य निर्धारण
इनपुट मूल्य$0.5 / 1M टोकन
आउटपुट मूल्य$2.5 / 1M टोकन
मिश्रित मूल्य (3:1)$1 / 1M टोकन
गति
कोई गति डेटा उपलब्ध नहीं
प्रदाता मूल्य रैंकिंग
प्रदाता मूल्य रैंकिंग
4 प्रदाता
सबसे सस्ता: ByteDanceसबसे महंगा: Kilo Gateway
प्रदाताइनपुटआउटपुट
1ByteDanceप्राथमिक
$0.5
$2.5
2NanoGPT
$0.5
$2.5
3OpenRouter
$0.5
$2.5
4Kilo Gateway
$0.5
$2.5
इस मॉडल के लिए विभिन्न API प्रदाताओं के मूल्य निर्धारण की तुलना करें।