Try Seedance 2.5 Uncensored Video from ByteDance →
Models
Agents
Workflows
Studio
PricingBlogDocs
ExploreDiscover models by categoryBrowse All ModelsBrowse the complete catalogSee FavoritesSign in to view saved models
Generative Media AgentCreate and edit media by chattingWorkflow AgentBuild visual workflows with Agent
OverviewThe platform at a glanceLearnSkills, knowledge, guardrailsAnatomyWhat makes agents reasonBuild Your AgentPick skills, set tier, deploy
Pre-built AgentsBrowse the catalog
Agent Usecases
Ad Campaign ManagerApp Event ManagerApp Review RepliesBarber BookingCustomer Win-BackEcommerce ListingsRestaurant Reviews
Sign InStart Building

Task History

Click to see output list

No tasks yet

Go to Models
Explore models/
LLM & ChatActive

ByteDance / seed-v2-pro

seed-v2-pro

bybytedance

Seed V2 Pro is ByteDance’s Seed2.0 Pro-class model for long-horizon reasoning, agent-style task planning, and image-grounded Q&A in chat workflows. (seed.bytedance.com)

Partner LLM
Model ID
seed-v2-pro
Provider
bytedance
Updated
1786973935
seed-v2-pro
11
Comments
Average rating : 4.5 (19 users)
Providerbytedance
Modelseed-v2-pro
Partner LLM
wiro playground—bytedance/seed-v2-pro
Reset to defaults
Delete All
0 / 20
Maximum 20 image allowed
Drop image to upload

OR

Click to browse your device

Supports: JPG, JPEG, PNG, GIF, WEBP, HEIC

Optional. Multiple image files separated by semicolon for visual understanding.

Required.

Required for chat history. Numeric or string ID.

Required for chat history. Numeric or string ID.

Sample outputs

No samples yet

Run this model to create outputs and build up samples.

Updated 1786973935

Overview

Seed V2 Pro is a multimodal large language model from ByteDance’s Seed team. (seed.bytedance.com) It’s built for complex reasoning, long-horizon instructions, and agent-style task execution. (seed.bytedance.com) You provide a question plus optional images for grounding. The model returns a chat-style text answer you can use in apps, tools, or support flows. (fal.ai) When thinking is enabled, some deployments can return a separate reasoning channel alongside the final answer. (fal.ai)

What you can build

  • Visual Q&A for screenshots, photos, and product images
  • Document help using captured pages, charts, and tables as images
  • Multi-step plans and checklists for operations, research, and analysis
  • Coding assistance for debugging, refactors, and explanations
  • Customer support drafts with strict tone and policy instructions
  • “Agent” style task breakdowns that track constraints and dependencies (seed.bytedance.com)

Inputs

  • A required instruction or question in plain text.
  • Optional image files for visual understanding. You can attach multiple images in one request when you need cross-image comparison.
  • Optional user and session identifiers (string or number) to keep multi-turn chat history consistent across runs.
  • Optional system-level instructions that set role, tone, and rules for the whole conversation.
  • Optional creativity controls that tune how deterministic or varied the wording is.
  • Optional repetition controls that reduce repeated phrases or encourage topic changes.
  • Optional thinking controls that let you force “think more” behavior, disable it, or let the model decide. Reasoning depth can be tuned from minimal to deep. (fal.ai)
  • Optional maximum output token budget to cap response length. Some deployments count both reasoning and final text against this limit. (live-turing-docs.cn.llm.tcljd.com)

Outputs

  • A text response suitable for chat UIs, reports, and tool instructions. (fal.ai)
  • Optional reasoning text returned separately when thinking is enabled (availability depends on the serving API). (fal.ai)
  • Optional updated conversation message history that you can pass back to continue the same thread. (fal.ai)

Limitations

  • Image understanding depends on input quality. Blurry photos, low resolution, heavy compression, or tiny text can cause missed details.
  • Long tasks can hit token limits. Deep thinking can consume part of the same output budget as the final answer. (live-turing-docs.cn.llm.tcljd.com)
  • Parameter support can vary by provider. Some deployments restrict sampling or penalty controls in reasoning modes. (docs.evolink.ai)
  • The model can still hallucinate or misread visuals. Validate high-stakes decisions with primary sources or human review. (seed.bytedance.com)

Safety & compliance

  • Don’t submit personal data you don’t have permission to share.
  • Don’t request instructions that enable wrongdoing, malware, or evasion.
  • For copyrighted images and documents, confirm you have rights to upload and analyze them.
  • Review outputs before using them in medical, legal, or financial decisions.

API quick start

Run seed-v2-pro with a single API call.

POST https://api.wiro.ai/v1/Run/ByteDance/seed-v2-pro
{
  "prompt": "Tell me about Türkiye",
  "inputImage": "https://your-cdn.com/input.png",
  "userId": "...",
  "sessionId": "..."
}
View full API docs

Discover, test, and run AI models, build workflows and agents with one unified API.

All systems operational
WiroAboutBlogCareersContact
ProductModelsAgentsPricingPartnerChangelogStatusFAQ
Getting StartedIntroductionAuthenticationProjectsCode ExamplesWiro MCP ServerSelf-Hosted MCPn8n IntegrationLLMs.txt
API ReferenceModelsRun a ModelModel ParametersTasksLLM & Chat StreamingWebSocketRealtime VoiceFiles
© 2026 Wiro AI. All rights reserved.
PrivacyTermsData Deletion