Try MiniMax H3 (Text-to-Video) (Image-to-Video) from MiniMax →
Models
Agents
WorkflowsStudioPricingBlogDocs
ExploreDiscover models by categoryBrowse All ModelsBrowse the complete catalogSee FavoritesSign in to view saved models
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

openai / gpt-5.4

gpt-5.4

byopenai

OpenAI GPT-5.4 takes text plus optional images and returns detailed answers, code, and extracted facts. It supports a 1,050,000-token context window.

Partner LLM
Model ID
gpt-5.4
Provider
openai
Updated
1781712398
gpt-5.4
2
Comments
Average rating : 5 (3 users)
Provideropenai
Modelgpt-5.4
Partner LLM
wiro playground—openai/gpt-5.4
Reset to defaults
Delete All
0 / 10
Maximum 10 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. GPT-5.4 only supports images.

Required.

Random

Required for chat history. Numeric or string ID.

Random

Required for chat history. Numeric or string ID.

Sample outputs

No samples yet

Run this model to create outputs and build up samples.

Updated 1781712398
## Overview GPT-5.4 by OpenAI is a long-context reasoning model that reads text and optional images, then writes a text response. It supports both single-turn prompts and multi-turn chat messages. You can dial up reasoning depth when you need multi-step analysis. This helps you turn screenshots, charts, and documents into clear explanations or extracted details. ## What you can build - Visual Q&A over screenshots, photos, and UI mockups - Chart and table reading, including extracting key numbers into clean text - Document review from images, like scanned pages or slide screenshots - Image-based troubleshooting for error dialogs, logs, and configuration screens - Code generation and refactors, guided by a spec and optional screenshots - Step-by-step workflows that require longer context and careful tracking ## Inputs - A required instruction written in plain text that explains the task and the expected format. - Optional chat history as a JSON-formatted list of messages (for example: user and assistant turns). When you provide message history, the single-turn prompt and system instructions are ignored. - Optional system-level instructions in plain text that set rules for tone, formatting, and boundaries. - Optional one or more image files for analysis. Supported formats include PNG, JPEG/JPG, WEBP, and non-animated GIF. - Optional image batching: you can include many images in one request, up to 500 images total, with a total payload up to 50 MB. - Optional end-user and session identifiers (string or number) to support chat history and session continuity. - Optional reasoning depth selection from fast to maximum-depth. Higher reasoning depth can use more internal reasoning tokens and may need a higher output token budget. - Optional verbosity control to keep responses short, balanced, or detailed. - Optional maximum output length, expressed as a token budget for the generated response. - Optional web search toggle to allow the model to browse the web when your prompt calls for it. ## Outputs The model returns generated text as an ordered list of text chunks. Join the chunks to form the final response. The text can include structured formatting like headings, bullet lists, tables in Markdown, or code blocks, depending on your instructions. ## Recommended settings - Use fast or low reasoning depth for short tasks like captioning a screenshot. - Use high or maximum reasoning depth for multi-step tasks like reconciling values across charts. - If you choose high reasoning depth, increase the output token budget to avoid empty or cut-off replies. - Set low verbosity for concise extraction and high verbosity for explanations and step-by-step plans. ## Limitations - GPT-5.4 accepts image input but does not support audio or video inputs. - The maximum context window is 1,050,000 tokens, and the maximum output is 128,000 tokens. - The model knowledge cutoff is August 31, 2025, so it may miss newer facts without web search. - Vision quality depends on the input image. Blurry screenshots, tiny text, glare, or heavy compression can cause missed details. - Images with watermarks or logos can be rejected. Images must also be clear enough for a human to understand. - Images with text in non-Latin alphabets may perform worse than Latin-script text. ## Safety & compliance OpenAI trains GPT-5.4 Thinking with policies that target disallowed content categories like harassment, sexual content, hate, self-harm, and wrongdoing. For computer-use scenarios, OpenAI uses a confirmation policy approach that can require extra confirmation for higher-risk actions. OpenAI also applies added cyber safeguards, including classifier-based blocking and account-level monitoring for high-risk cyber misuse.

API quick start

Run gpt-5.4 with a single API call.

POST https://api.wiro.ai/v1/Run/openai/gpt-5.4
{
  "prompt": "Tell me about Türkiye",
  "inputImage": "https://your-cdn.com/input.png",
  "user_id": "...",
  "session_id": "..."
}
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