GPT 5

Supported Functionality

Item	Specification
Input	Text, image
Output	Text
Context	400,000 tokens
Max Output	128,000 tokens
Vision	✓ Supported
Function Calling	✓ Supported

Description

GPT-5 is OpenAI's general reasoning model for cross-domain coding, reasoning, and agentic tasks. OpenAI provides the gpt-5 alias and the dated gpt-5-2025-08-07 snapshot, indicating an August 7, 2025 version snapshot; its knowledge cutoff is September 30, 2024. It accepts text and image inputs, produces text, and offers a 400,000-token context window with up to 128,000 output tokens. OpenAI has not disclosed its parameter count or detailed architecture.

A key change is the integration of configurable reasoning and tool use into a general workflow. The model supports minimal, low, medium, and high reasoning effort, allowing applications to balance speed, token cost, and analytical depth. Streaming, function calling, and Structured Outputs support composable agentic systems. As of August 2026, OpenAI classifies it as a previous-generation model and recommends GPT-5.6 for new projects.

Key Capabilities

• Adjustable Deep Reasoning: Four reasoning-effort levels cover quick questions, rigorous analysis, mathematical derivation, and multi-constraint decisions.
• Software Development: Generates, explains, debugs, and refactors code while using repository context for cross-file engineering problems.
• Agentic Tool Use: Function calling connects search, databases, business APIs, or execution environments for multi-step information processing and actions.
• Vision Understanding: Analyzes screenshots, charts, document pages, and UI designs for visual question answering, extraction, and frontend assistance.
• Long-Context Processing: A 400,000-token window accommodates large document sets, repositories, and conversations, although critical workloads should evaluate long-range recall.
• Structured Generation: Structured Outputs return schema-constrained results that applications can validate, store, and pass to subsequent program steps.
• Multilingual Content Processing: Handles cross-language question answering, translation, summarization, and professional writing while incorporating reasoning and tool results.

Technical Strengths

Feature	Benefit
Unified Reasoning Model	One model handles analysis, coding, vision, and tool use, reducing the complexity of routing across specialized models.
Four Reasoning Levels	Developers can control latency and cost for routine requests while allocating deeper reasoning to difficult tasks.
Very Large Input Capacity	The 400K-token window brings more code and documentation into one request for cross-section and cross-file synthesis.
Large Output Capacity	A 128K-token ceiling accommodates long reports, detailed plans, and large code results with less truncation risk.
Vision-Text Fusion	Images can share context with instructions, data, and code for work involving charts, screenshots, and document pages.
Reliable System Integration	Function calling, streaming, and Structured Outputs support parseable, executable applications with progressive result delivery.