Introducing the Q4 API: Saudi market data, every figure sourced, in your own productsGet your API key
Q4

Q4 Terminal

AI for Saudi company research

GET /v1/quote/2222
{
  "symbol": "2222",
  "price": 25.42,
  "change_percent": 0.71
}

Q4 API

Saudi market data

Q4 MCP

Q4 data inside your AI assistant

Investment teams

  • Equity research

    Verify every number

  • Asset managers

    After the call, before the memo

  • Corporate finance and advisory

    Benchmark peers on verified figures

Companies and builders

  • Investor relations

    Know how your peers report

  • AI and fintech builders

    Saudi market data in your AI tools

  • About

    Why Q4 exists and how we build it

  • Blog

    Notes on Saudi market research

Sign inBook a Demo
Blog

GPT-6 Sol and Luna now available in Q4

Product/23 September 2026/2 min read

We’re excited to bring OpenAI’s GPT-6 Sol and Luna to Q4.

GPT-6 with Q4
GPT-6 Sol and Luna are now available in Q4 chat.

GPT-6 is the generation after GPT-5.6, and its headline change is reliability rather than raw capability. OpenAI released its flagship, Astra, on 3 September, then Sol and Luna on 22 September as smaller, cheaper models that it says bring Astra’s improvements to everyday use. Sol and Luna are the two we have added to Q4.

OpenAI says GPT-6 Sol makes about half as many mistakes as GPT-5.6 Sol on its internal factuality evaluation, which is built from real conversations where users flagged errors, reaching the reliability of Astra. Sol costs half as much as GPT-5.6 Sol, and Luna’s input price drops from $0.20 to $0.10 per million tokens.

That error rate is the number that matters for research. In financial work a confident wrong figure is worse than no figure at all. Q4 already ties every number to the filing it came from, and a model that makes fewer mistakes means fewer answers that need correcting in the first place.

The two models map onto how Q4 gets used. OpenAI positions Sol for complex work and Luna for high-volume tasks with a clear goal, like summarizing documents and extracting information. In Q4, Sol counts toward your daily model allowance, so save it for the question you need to get right. Luna has no daily limit, which suits pulling the same figure out of dozens of filings.

To try them, open Q4 chat, choose ChatGPT in the model picker and select GPT-6 Sol or GPT-6 Luna. Both are available in Q4 now.

Frequently asked questions

How is GPT-6 different from GPT-5.6?

The main change is reliability. OpenAI says GPT-6 Sol makes about half as many mistakes as GPT-5.6 Sol, reaching the reliability of its flagship, GPT-6 Astra. Sol and Luna also cost half or less of their 5.6 versions.

Which GPT-6 models are available in Q4?

GPT-6 Sol and GPT-6 Luna, in the ChatGPT group of the Q4 chat model picker. OpenAI’s flagship, GPT-6 Astra, is not currently offered in Q4.

Is there a usage limit on GPT-6 in Q4?

GPT-6 Sol counts toward a daily model allowance. GPT-6 Luna, GPT-5.6 Luna and Auto have no daily limit.

Do GPT-6 answers in Q4 include citations?

Yes. GPT-6 Sol and Luna run with Q4’s research tools, so answers draw on Tadawul filings, earnings calls and company data with citations to the source.

More from Q4

11 September 2026Product
Q4 connected to Claude, ChatGPT, and other AI tools through MCP

What Q4 MCP brings to your AI workflow

Connect Claude, ChatGPT or Cursor to Q4 and your assistant can research Saudi listed companies from filings, earnings calls and financial data, with a citation on every number.

Read more

Bring cited Saudi research into the tools you already use.

Book a DemoSee Q4 MCP
Product
Q4 TerminalQ4 MCPQ4 APIDocsUse Cases
Company
AboutBlog
Legal
Privacy PolicyTerms of ServiceLegal
Contact
Contact salesLinkedInXcontact@q4.sa