OpenCode Go Visual Studio 2026

OpenCode Go.
In Visual Studio.

Bring your OpenCode Go or Go Plus subscription to your C# and .NET workspace. Connect with your API key, choose your coding models, and keep your remaining allowance in view.

BRING YOUR OWN SUBSCRIPTION Your OpenCode account. One API key.

Your requests use your Go subscription allowance. iolys connects directly to the Go service.

How subscription access works ↗

Get iolys Explore the integration ↓

OpenCode Go configuration in iolys for Visual Studio 2026, showing a masked API key and available models with capability badges and selection checkboxes. — Choose the models you want close to your code.Enlarge ↗

Product UI rendered with demonstration data. The captures show no live conversation or account. Model availability and allowances depend on your OpenCode account.

01 / Your model shortlist

Several model families.
One place to choose.

Keep the Go models you use for implementation, code review, and debugging in your chat picker. Available families include GPT Luna, DeepSeek, GLM, Kimi, MiniMax, and Qwen, following OpenCode’s current catalog.

YOUR SELECTION

Keep your picker focused.

Choose which models appear in chat from the Go connection panel. Select the active model beside your prompt whenever you start a task.

MODEL CAPABILITIES

Choose the support your task needs.

Use the catalog’s capability badges and known context limits to guide your choice. Compatible models can use iolys tools for your solution and accept image attachments.

02 / Reasoning effort

Pick your model.
Set its reasoning effort.

Choose the reasoning effort in the chat model picker. The available levels follow the model you select.

GPT-6 Luna and GPT-5.6 LunaNone, Low, Medium, High, and Xhigh.
DeepSeek V4 and GLM-5.2DeepSeek V4 offers Low, Medium, High, and Max. GLM-5.2 offers High and Max.
MiniMax M3Choose None or Thinking. Thinking is selected initially for adaptive reasoning.

The initial effort is High for the other listed models. Models without an effort selector use their server-side reasoning behavior.

Demonstration chat with GPT-6 Luna selected and the None, Low, Medium, High, and Xhigh reasoning choices expanded. — Reasoning controls beside your prompt.

03 / Context window

See what fills
your model's context.

Open the context gauge to see how much of the selected model's known context window is in use, with a breakdown of messages, tools, IDE state, and files.

The latest model call.The total uses reported input tokens when available. Otherwise, iolys estimates the active conversation.
A breakdown you can inspect.Category counts are estimated locally. An unknown model limit remains unavailable.

Context usage measures the input sent to the model. Your subscription allowance is tracked separately in Usage.

Example context window at 12 percent, with expanded System and User context categories for messages, tools, IDE state, and files. — Context usage and its estimated breakdown.

04 / Account usage

See your allowance.
Know when it resets.

Open Usage from the chat to read the consumption reported for your OpenCode account, without leaving Visual Studio.

Three usage windows.Check five-hour, weekly, and monthly consumption and remaining percentages.
Reset times and reached limits.See when each reported allowance renews and whether a limit has been reached.
Your whole account.These counters include requests from other clients using the account, alongside your work in iolys.

Only the counters reported by OpenCode are shown. The usage endpoint does not report a monetary balance or identify whether your subscription is Go or Go Plus.

Example OpenCode Go Usage window displaying five-hour, weekly, and monthly account consumption with remaining percentages and reset times. — Account consumption and resets. Example values.

Connect OpenCode Go

One API key. Your Go subscription in Visual Studio.

Connect your OpenCode Go or Go Plus account, choose your models, and start your next C# or .NET task.

01
Sign in to the OpenCode Console, subscribe to Go or Go Plus, and copy your API key.
02
Open Manage Models in iolys, choose OpenCode Go, paste your key, and run Test Connection. iolys configures the Go endpoint automatically.
03
Choose the models to show in the chat picker, then select a model and a supported reasoning effort for your Visual Studio conversation.

OpenCode Go FAQ

OpenCode Go and Visual Studio 2026.

Visit the official OpenCode Go website ↗

How can I use OpenCode Go in Visual Studio for C#?

Install the iolys extension for Visual Studio 2026 and open your C# solution. Sign in to the OpenCode Console, subscribe to Go or Go Plus, and copy your API key. Open Manage Models in iolys, choose OpenCode Go, paste your key, and run Test Connection. iolys configures the Go endpoint automatically. Choose the models to show in the chat picker, then select a model and a supported reasoning effort for your Visual Studio conversation. You can then ask OpenCode Go to help with your C# and .NET code from the iolys workspace.

Is there a VSIX extension to use OpenCode Go in Visual Studio?

Yes. iolys is a VSIX extension for Visual Studio 2026, and OpenCode Go is one of the provider connections it ships with. Install the iolys VSIX once from the Visual Studio Marketplace, then add OpenCode Go from the Manage Providers panel; there is no separate VSIX per provider.

Can I use my OpenCode Go subscription in Visual Studio?

Yes. Connect your OpenCode Go or Go Plus API key to the dedicated OpenCode Go provider in iolys. Usage counts toward your OpenCode account's allowance. No OpenCode CLI installation is required for this connection.

Is OpenCode Go the same as OpenCode Zen?

No. Go is OpenCode's subscription service, while Zen is its separate pay-as-you-go service. The dedicated iolys provider uses Go's catalog and endpoints. OpenCode manages plan limits and any account-level option to use a balance beyond those limits.

Where do I select models and reasoning effort?

Use Manage Models to configure your API key and the models visible in chat. Select the active model and its reasoning effort in the chat model picker. Test Connection refreshes the available catalog.

Which reasoning levels can I choose?

GPT-6 Luna and GPT-5.6 Luna offer none, low, medium, high, and xhigh; DeepSeek V4 offers low, medium, high, and max; GLM-5.2 offers high and max; MiniMax M3 offers none and thinking. Other models keep their server-side reasoning behavior without an effort selector. Available choices depend on the selected model.

Can I use images and iolys tools with OpenCode Go?

Yes, when the selected model advertises the required capabilities. Models with image input support can receive image attachments, and tool-capable models can use the tools supplied by iolys for your solution. Capabilities and known context limits are shown from the model catalog.

What does OpenCode Go Usage show?

The Usage window shows the five-hour, weekly, and monthly percentages, remaining percentages, reset times, and reached limits reported for your account. It includes usage from other clients. Only the counters reported by OpenCode are shown; the endpoint does not report a monetary balance or identify whether your plan is Go or Go Plus.