OpenCode Qwen3.7-Plus Supported Model Docs (2026)

OpenCode Qwen3.7-Plus supported model docs confirm that qwen3.7-plus is available through QwenCloud’s OpenCode integration, including Token Plan, Coding Plan, and pay-as-you-go configurations. It is also listed for the opencode-go provider, where the documented model supports text and image input, reasoning, a 1,000,000-token context window, and up to 65,536 output tokens.

The practical answer is: configure the provider you actually pay for, use the exact model ID qwen3.7-plus, restart OpenCode, then run /models and select QwenCloud or OpenCode Go. Do not assume that seeing Qwen3.7-Plus in one provider’s catalog means it is automatically enabled for every API key or subscription.

Is Qwen3.7-Plus supported in OpenCode?

Yes. QwenCloud’s OpenCode documentation explicitly includes qwen3.7-plus in its example model maps. The model appears under Token Plan Personal Edition, Token Plan Team Edition, Coding Plan, and pay-as-you-go examples. That matters because “supported” has two layers: OpenCode must be able to speak the provider’s API, and your chosen QwenCloud billing product must expose the model to the credentials you use.

OpenCode itself is an open-source coding agent with terminal and desktop interfaces. It can be connected to different model providers, so support is generally provider-scoped rather than a promise that every installation receives a model without configuration. In this case, the QwenCloud documentation supplies a working provider definition, endpoint, model ID, and thinking configuration for Qwen3.7-Plus.

There is also an OpenCode Go listing for the model. Its provider identifier is opencode-go, its API format is Anthropic Messages, and its base URL is https://opencode.ai/zen/go. That is a separate route from bringing your own QwenCloud key. Treat OpenCode Go and direct QwenCloud access as distinct products with distinct authentication, billing, quotas, and possibly different model metadata.

What is the Qwen3.7-Plus model ID in OpenCode?

Use qwen3.7-plus. Keep the lowercase spelling and punctuation exactly as shown. The display name may be “Qwen3.7 Plus,” but OpenCode provider configuration uses the machine-facing ID. A common failure mode is using a display label, an older Qwen alias, or a provider-specific ID from another tool and receiving a model-not-found error.

For OpenCode Go, the fully qualified provider/model reference is commonly represented as opencode-go/qwen3.7-plus. For a manually configured QwenCloud provider, the provider key is the name you assign in opencode.json—for example, bailian-coding-plan—while the nested model key remains qwen3.7-plus.

How do you configure Qwen3.7-Plus in OpenCode?

First, install or update OpenCode. QwenCloud’s guide specifies Node.js 18 or later and uses the opencode-ai package. Then create or edit the OpenCode configuration file: ~/.config/opencode/opencode.json on macOS and Linux, or C:\Users\<Your-Username>\.config\opencode\opencode.json on Windows. The exact provider block depends on whether you use a QwenCloud Coding Plan, Token Plan, or pay-as-you-go billing.

For a QwenCloud Coding Plan, the provider uses the Anthropic SDK adapter and QwenCloud’s Coding Plan endpoint. This is the smallest useful configuration for Qwen3.7-Plus:

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "bailian-coding-plan": {
      "npm": "@ai-sdk/anthropic",
      "name": "QwenCloud",
      "options": {
        "baseURL": "https://coding-intl.dashscope.aliyuncs.com/apps/anthropic/v1",
        "apiKey": "YOUR_API_KEY"
      },
      "models": {
        "qwen3.7-plus": {
          "name": "Qwen3.7 Plus",
          "options": {
            "thinking": {
              "type": "enabled",
              "budgetTokens": 1024
            }
          }
        }
      }
    }
  }
}

Replace YOUR_API_KEY with the dedicated key for that billing product—not a generic key copied from a different Qwen account or region. The Coding Plan example assigns a 1,024-token thinking budget. QwenCloud’s Token Plan and pay-as-you-go examples also include Qwen3.7-Plus, but show an 8,192-token thinking budget instead. Copy the block belonging to your plan rather than mixing the Coding Plan endpoint with a pay-as-you-go key.

How do you use Qwen3.7-Plus with QwenCloud pay-as-you-go?

The pay-as-you-go example changes both the adapter and endpoint. It uses @ai-sdk/openai-compatible with QwenCloud’s OpenAI-compatible endpoint, while retaining qwen3.7-plus as the model key. This distinction is important: model support is not just a string in a menu. The adapter determines request formatting, streaming behavior, and how OpenCode translates its agent requests into the provider’s API.

{
  "$schema": "https://opencode.ai/config.json",
  "provider": {
    "bailian-payg": {
      "npm": "@ai-sdk/openai-compatible",
      "name": "QwenCloud",
      "options": {
        "baseURL": "https://dashscope-intl.aliyuncs.com/compatible-mode/v1",
        "apiKey": "YOUR_API_KEY"
      },
      "models": {
        "qwen3.7-plus": {
          "name": "Qwen3.7 Plus",
          "options": {
            "thinking": {
              "type": "enabled",
              "budgetTokens": 8192
            }
          }
        }
      }
    }
  }
}

If you already have a larger provider block, add the Qwen3.7-Plus model object inside its existing models map instead of creating duplicate provider names. JSON does not tolerate duplicate keys in a useful, portable way: one definition can silently override another depending on the parser.

How do you verify that Qwen3.7-Plus is available?

Save the file, fully restart OpenCode, type /models, search for “QwenCloud,” and choose Qwen3.7 Plus. This is the verification sequence documented by QwenCloud, and it is preferable to guessing from a stale model picker. If the model does not appear, inspect configuration validation output first, then verify that the API key belongs to the plan named by your provider configuration.

1. Run opencode -v to confirm that OpenCode is installed and callable.
2. Validate that opencode.json is in the correct OS-specific configuration directory.
3. Confirm the provider endpoint and adapter match the billing plan tied to your API key.
4. Restart OpenCode after saving changes; do not rely on a previously open session.
5. Run /models, search for the provider, and select qwen3.7-plus.
6. Send a small prompt before giving the agent a repository-wide task, so authentication and model routing failures are cheap to diagnose.

What capabilities does Qwen3.7-Plus expose in OpenCode?

The OpenCode Go model listing documents text and image input, text output, reasoning support, a 1,000,000-token context window, and a 65,536-token maximum output. Those numbers describe the provider’s published model metadata, not a guarantee that every agent turn will preserve a million tokens of useful repository context. An agent also spends context on its instructions, tool schemas, file contents, shell output, and conversation history.

QwenCloud’s examples explicitly enable thinking for Qwen3.7-Plus. Thinking budget is not simply “better quality = larger number”: it is a cap that changes how much room the model has for internal reasoning in a request. Start with the documented default for your plan. Raise it only after measuring whether the added latency and token use improve the coding tasks you actually run—such as multi-file changes, debugging, or test repair.

Why is Qwen3.7-Plus missing from my OpenCode model list?

The most likely causes are a mismatched endpoint, an API key for the wrong QwenCloud product, a typo in the model key, or a configuration file that OpenCode is not loading. Another possibility is that the provider has changed its entitled model list. A model page, blog post, or another developer’s config is evidence that a model exists; it is not proof that your credentials can invoke it today.

Avoid solving this by immediately hardcoding arbitrary context-window or token-limit overrides. Overrides can be useful when provider metadata is wrong, but they cannot grant API access and may make OpenCode behave poorly if they overstate an actual limit. Establish a successful minimal call first. Then tune model metadata only when you have a specific, reproducible reason.

Should you use Qwen3.7-Plus through OpenCode Go or QwenCloud directly?

Choose based on where you want the provider relationship to live. OpenCode Go gives you an OpenCode-hosted provider route with its own model record and pricing metadata. Direct QwenCloud configuration gives you a QwenCloud plan or pay-as-you-go key and makes the endpoint and model definition explicit in your config. Neither route is universally better: the meaningful differences are account ownership, billing, quota policy, regional availability, and the operational convenience you want.

For teams, make the choice deliberately and commit a sanitized configuration template to internal documentation—not live keys. Record the model ID, provider key, endpoint type, and expected thinking settings. That turns an individual developer’s working setup into a repeatable development environment.

Want Qwen3.7-Plus in a different coding-agent workflow?

Cline is an open-source coding agent that runs in an IDE, terminal, and SDK, with project-wide edits, terminal execution, planning, rules, MCP/plugin extension, and automation support. Its site says developers can use their own keys, endpoints, or weights, including OpenAI-compatible APIs—useful if your question about OpenCode support is really about retaining flexibility in how you access Qwen3.7-Plus.

For a managed route, ClinePass includes Qwen3.7-Plus as a balanced model for regular feature work, debugging, and code review. At publication, Cline lists a promotional first month at $4.99 and then $9.99 per month, with a note that additional processing fees may apply; bring-your-own-key workflows remain available separately. That is worth considering if you want an agent harness around the model without separately managing provider setup and API keys.