Skip to content

DISCOVERY · MODELS

Syniray Models

The right model for every question.

15 models — Syniray 2.0 and models from 7 AI companies — in one place. See what each one offers, which plan includes it and where you can use it.

Explore the plans

Choosing

Choose per message, or let Syniray choose

  • Per message

    Choose the model for each message in the Chat, for each question in the Academy and for each session in Syniray Code.

  • Best Mode

    On Pro and Max, Best Mode chooses the most suitable model for each question for you.

  • Where they are available

    Every model of your plan is available in the Chat, the Academy and Syniray Code's cloud. Deep Research and Fact Check run on a selection of them.

The catalogue

Every model, by plan

Grouped by the plan that first includes them. Each card shows who makes the model, whether it reads images, its reasoning levels and where else you can use it.

Every plan

Included in Free, Pro and Max.

  • Syniray 2.0

    by Syniray

    Free+
    • Reads images

    Reasoning levels

    • Low
    • High
    • Max

Pro and Max

Included from the Pro plan.

  • GPT-6.1 Terra

    by OpenAI

    Pro+
    • Reads images
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High
    • Very high
    • Max
  • Gemini 3.8 Flash

    by Google DeepMind

    Pro+
    • Reads images
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High
  • Claude Sonnet 5.5

    by Anthropic

    Pro+
    • Reads images
    • Deep Research and Fact Check
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High
    • Very high
    • Max
  • DeepSeek V4 Pro

    by DeepSeek AI

    Pro+
    • Text only
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • High
    • Max
  • Kimi K3.1 Pro

    by Moonshot AI

    Pro+
    • Reads images
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • High
    • Max
  • GLM-5.3

    by Z.AI

    Pro+
    • Text only
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • High
    • Max
  • GPT-6.1 Sol

    by OpenAI

    Pro+
    • Reads images
    • Deep Research and Fact Check
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High
    • Very high
    • Max
  • Claude Opus 5.5

    by Anthropic

    Pro+
    • Reads images
    • Deep Research and Fact Check
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High
    • Very high
    • Max
  • Grok 4.7

    by xAI

    Pro+
    • Reads images
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High
    • Very high
  • GPT-6 Luna

    by OpenAI

    Pro+
    • Reads images
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High
    • Very high
    • Max
  • Claude Haiku 5.5

    by Anthropic

    Pro+
    • Reads images
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High
    • Very high
    • Max
  • Gemini 4 Argon

    by Google DeepMind

    Pro+
    • Reads images
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High

Max only

The most capable models, included only in Max.

  • GPT-6 Astra

    by OpenAI

    Max
    • Reads images
    • Deep Research and Fact Check
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High
    • Very high
    • Max
  • Claude Fable 5.1

    by Anthropic

    Max
    • Reads images
    • Deep Research and Fact Check
    • Also on the Enterprise API

    Reasoning levels

    • Low
    • Medium
    • High
    • Very high
    • Max

Reasoning effort

How long a model thinks before answering

Models that reason can work through a question for longer before they answer. Each model has its own levels. On Pro, the level is chosen automatically for each question. On Max, you can choose it yourself in the models and products that support it.

In Syniray Code

Your own key and local models

  • Your own API key

    In Syniray Code you can connect a key from OpenAI, Anthropic, Google, DeepSeek or an OpenAI-compatible endpoint, billed by your provider.

  • Local models

    When Syniray Code runs on your computer, it can use models served by Ollama or by a local OpenAI-compatible server.

Questions

Frequently asked questions

How many models are there?

15: Syniray 2.0 and models from 7 AI companies. Free includes Syniray 2.0; Pro adds the Pro models; Max adds two Max-only models.

Which model should I use?

It depends on the question and on how much you want to spend. If you are not sure, Best Mode (Pro and Max) chooses for you; Syniray 2.0 is available on every plan.

Why do some models use more credits?

Each answer is charged in credits for the text the model reads and writes, and each model uses credits in proportion to what it costs to run — so the most capable models use more credits for the same answer. Your balance and what each use cost are in Settings › Usage.

What does “reads images” mean?

That the model can analyse images you attach. Syniray only lets you attach images to models that read them.

Can I choose the reasoning level?

On Max, in the models and products that support it. On Pro the level is set automatically for each question.

Can I use my own API key or local models?

Yes, in Syniray Code: your own key from OpenAI, Anthropic, Google, DeepSeek or an OpenAI-compatible endpoint, or local models on your computer.

Are these models available through an API?

Syniray Enterprise offers 14 of them through an OpenAI-compatible API. Syniray 2.0 is not part of it.

What is Syniray 2.0?

Syniray 2.0 is the model Syniray offers by default, built on open-source model technology. It is the model of the Free plan and is available on every plan; it reads the images you attach on Pro and Max, and searches the web when a question needs it.

Use every model in one place

Free includes Syniray 2.0; Pro and Max open the rest of the catalogue.

Explore the plans