GPT 6.2 Sol by OpenAI
GPT-6.2 Sol by OpenAI is a long-context text model with image understanding. It writes code, drafts documents, and supports tool-based agent workflows.
## Overview
GPT-6.2 Sol is OpenAI's general-purpose model for everyday professional work, from coding to powering agents at scale. It takes text and optional images and returns text. With a 1,050,000-token context window and up to 128,000 output tokens, it handles long, multi-step tasks, and one model can write, code, reason and coordinate tools in a single workflow.
## What you can build
- A coding assistant that reviews diffs, writes patches and explains changes in plain language
- A document drafting tool for specs, SOPs, reports and internal memos
- A vision Q&A helper that explains screenshots, charts, UI states or product photos
- An agent loop that calls your functions, waits for tool results, then continues the task
- A chat experience with per-user and per-session history
## Inputs
- Your request as a plain-text prompt, or a conversation history as a JSON array of messages with roles and content.
- Up to 10 input images per run: PNG, JPEG, WEBP or non-animated GIF.
- Optional system instructions, up to 5,000 characters.
- Optional user and session IDs to keep chat history consistent.
- A reasoning effort level: low, medium, high, xhigh or max. Reasoning can't be turned off.
- A verbosity level that controls how short or detailed the response is.
- An optional maximum output length, up to 128,000 tokens.
- Optional tool definitions as JSON to enable function calling, with a tool choice (auto, none or required) and an option to allow several tool calls in one turn.
- An optional web search toggle for fresh information. Each search costs $0.01.
- To resume a tool workflow, the prior run's task token and the tool results as a JSON array.
## Outputs
- A text response: prose or code, depending on your instructions.
- Tool-call requests when tool calling is enabled. After you send the tool results back, the model continues and produces the final answer.
## Recommended settings
- Medium reasoning effort for most work.
- Low reasoning effort for short rewrites, classification or fast replies.
- High or xhigh reasoning effort for hard debugging and multi-step planning.
## Limitations
- Output is text only. Audio and video input are not supported.
- The context window is 1,050,000 tokens, with up to 922,000 input tokens. Requests over 272K input tokens are billed at higher long-context rates.
- The knowledge cutoff is June 1, 2026. Turn on web search for later events.
- Blurry screenshots, heavy compression, tiny text, low-quality scans and messy layouts reduce accuracy.
## Safety & compliance
- OpenAI applies safety filtering, and some high-risk cybersecurity, biological or chemical requests are refused.
- Don't use it to generate or improve malicious code, instructions for harm, or prohibited content.
- When you enable tool use, treat tool outputs as untrusted input and validate them before acting on them, especially in security-sensitive workflows.
API quick start
Run gpt 6.2 sol with a single API call.
{
"prompt": "How to roast coffee beans at home?",
"inputImage": "https://your-cdn.com/input.png",
"user_id": "...",
"session_id": "..."
}curl -X POST "https://api.wiro.ai/v1/Run/openai/gpt-6-2-sol" \
-H "Content-Type: application/json" \
-H "x-api-key: YOUR_WIRO_API_KEY" \
--data-binary @- <<'JSON'
{
"prompt": "How to roast coffee beans at home?",
"inputImage": "https://your-cdn.com/input.png",
"user_id": "...",
"session_id": "..."
}
JSONOther models you might like
Explore similar llm & chat models.



