Seed 2.1 Turbo

Name: Seed 2.1 Turbo
Author: ByteDance

Multimodal

ByteDance

ByteDance's low-cost, low-latency next-generation agent model for large-scale production.

Key Specifications

Parameters

Context

Release Date

June 24, 2026

Average Score

61.0%

API Documentation Results Blog

Timeline

Key dates in the model's history

Announcement

June 24, 2026

Last Update

August 29, 2026

Today

October 22, 2026

Technical Specifications

Parameters

Training Tokens

Knowledge Cutoff

Family

Capabilities

MultimodalZeroEval

Benchmark Results

Model performance metrics across various tests and benchmarks

Multimodal

Working with images and visual data

MathVista

• Self-reported

90.5%

Other Tests

Specialized benchmarks

AetherCode

• Self-reported

67.9%

Agent Startup Bench

• Self-reported

54.0%

APEX-Agents

• Self-reported

29.2%

ArcAGI2

• Self-reported

61.3%

Artifacts Bench

• Self-reported

47.0%

BabyVision

• Self-reported

62.9%

Beyond AIME

• Self-reported

88.0%

BLINK

• Self-reported

79.4%

BrowseComp

With search • Self-reported

84.9%

ChartQAPro

• Self-reported

70.9%

CharXiv-D

• Self-reported

94.6%

CharXiv-R

With tools • Self-reported

83.6%

ClawEval-MM

Pass^3 • Self-reported

46.0%

ContPhy

• Self-reported

61.1%

CreativeWork

• Self-reported

34.5%

CrossVid

• Self-reported

63.2%

CyberGym

• Self-reported

67.0%

DeepSWE

• Self-reported

23.0%

Doubao Multi-Turn Bench

• Self-reported

49.0%

DUDE

• Self-reported

83.1%

DynaMath

• Self-reported

68.1%

EmbSpatialBench

• Self-reported

82.5%

EMMA

• Self-reported

78.4%

ERQA

• Self-reported

71.3%

Finance Agent v1.1

• Self-reported

56.0%

FrontierCS

• Self-reported

50.8%

FrontierScience Olympiad

• Self-reported

76.0%

FrontierScience Research

• Self-reported

33.3%

GameWorld

• Self-reported

25.9%

GDPval

• Self-reported

82.7%

HorizonMath

• Self-reported

2.0%

Humanity's Last Exam

Text-only, with search • Self-reported

54.6%

Image2FloorPlan

• Self-reported

35.9%

KINA

• Self-reported

46.6%

LiveMathematicianBench

• Self-reported

27.7%

LiveSports-3K

• Self-reported

77.1%

LongVideoBench

• Self-reported

80.6%

LVBench

• Self-reported

76.8%

MathArena Apex

• Self-reported

35.4%

MathVerse

Vision-Only • Self-reported

89.2%

MathVision

With tools • Self-reported

92.7%

MCP Atlas

• Self-reported

80.3%

MeasureBench

avg. real & synthetic • Self-reported

58.9%

Minerva

• Self-reported

65.9%

MMLongBench-128K

• Self-reported

76.9%

MMMU-Pro

With tools • Self-reported

82.2%

MMSIBench

Circular • Self-reported

31.4%

MobileWorld

• Self-reported

70.0%

MotionBench

• Self-reported

74.8%

MSQA

• Self-reported

42.0%

NL2Repo

• Self-reported

43.7%

OCRBench_V2

• Self-reported

62.8%

OfficeQA Pro

Multimodal • Self-reported

71.1%

OneMillion Bench

• Self-reported

66.6%

OSWorld

• Self-reported

76.4%

OVBench

• Self-reported

69.7%

OVOBench

• Self-reported

79.2%

PostTrainBench

• Self-reported

18.3%

PresentBench

• Self-reported

48.3%

Program Bench

• Self-reported

49.4%

RealWorldQA

• Self-reported

86.3%

Repo Env

• Self-reported

46.7%

SciCode

• Self-reported

57.8%

SeedClawBench

• Self-reported

63.8%

SimpleVQA

• Self-reported

71.1%

SuperChem

• Self-reported

56.6%

SuperGPQA

• Self-reported

67.4%

SWE-Atlas

• Self-reported

30.6%

SWE-Bench Pro

• Self-reported

57.0%

Terminal-Bench 2.1

• Self-reported

67.6%

TOMATO

• Self-reported

56.8%

Toolathlon

• Self-reported

49.1%

Trae Code Gen

JavaScript • Self-reported

59.7%

Trae Error Fix

Go • Self-reported

56.7%

TreeBench

• Self-reported

71.1%

TVBench

• Self-reported

77.2%

VideoHolmes

• Self-reported

67.6%

Video-MME

• Self-reported

89.0%

VideoSimpleQA

• Self-reported

71.4%

VisFactor

• Self-reported

43.9%

VisuLogic

• Self-reported

52.9%

VLMsAreBiased

• Self-reported

68.3%

Web Bench

• Self-reported

73.6%

WildClawBench

• Self-reported

62.8%

Workspace Bench

• Self-reported

54.7%

WorldBench

• Self-reported

63.7%

WorldVQA

• Self-reported

48.6%

xDailyBench

• Self-reported

56.4%

ZEROBench

• Self-reported

11.0%

License & Metadata

License

proprietary

Announcement Date

June 24, 2026

Last Updated

August 29, 2026

Similar Models

All Models

Seed 2.1 Pro

ByteDance

Released:Jun 2026

Seed 1.8

ByteDance

Best score:0.9 (MMLU)

Released:Feb 2026

Price:$0.25/1M tokens

Seed 2.0 Mini

ByteDance

Best score:0.8 (GPQA)

Released:Feb 2026

Price:$0.10/1M tokens

Seedance 1.5 Pro

ByteDance

Released:Jan 1970

Seed 2.0 Pro

ByteDance

Best score:0.9 (GPQA)

Released:Feb 2026

Price:$0.50/1M tokens

Seed 2.0 Lite

ByteDance

Best score:0.9 (GPQA)

Released:Feb 2026

Doubao Seed Code

ByteDance

Released:Mar 2026

GLM-5V-Turbo

Zhipu AI

Released:Apr 2026

Price:$1.20/1M tokens

Recommendations are based on similarity of characteristics: developer organization, multimodality, parameter size, and benchmark performance. Choose a model to compare or go to the full catalog to browse all available AI models.