From 77ca37797134999dd2838a45bf1d5f736ea61db6 Mon Sep 17 00:00:00 2001
From: "@taltas" <6816042+taltas@users.noreply.github.com>
Date: Fri, 14 Aug 2026 13:05:16 +0000
Subject: [PATCH 1/3] docs(provider): document OpenCode Go models

---
 docs/providers/index.json     |  6 +++++
 docs/providers/opencode-go.md | 48 +++++++++++++++++++++++++++++++++++
 2 files changed, 54 insertions(+)
 create mode 100644 docs/providers/opencode-go.md

diff --git a/docs/providers/index.json b/docs/providers/index.json
index 92c13d53..a1f52bbb 100644
--- a/docs/providers/index.json
+++ b/docs/providers/index.json
@@ -90,6 +90,12 @@
       "extension": true,
       "cloud": false
     },
+    {
+      "id": "providers/opencode-go",
+      "title": "OpenCode Go",
+      "extension": true,
+      "cloud": false
+    },
     {
       "id": "providers/openrouter",
       "title": "OpenRouter",
diff --git a/docs/providers/opencode-go.md b/docs/providers/opencode-go.md
new file mode 100644
index 00000000..78347139
--- /dev/null
+++ b/docs/providers/opencode-go.md
@@ -0,0 +1,48 @@
+---
+sidebar_label: OpenCode Go
+description: Use OpenCode Go with Zoo Code for curated open coding models through a low-cost subscription.
+keywords:
+  - opencode go
+  - glm-5.3
+  - qwen3.8 max
+  - coding models
+  - zoo code
+  - api provider
+---
+
+# Using OpenCode Go With Zoo Code
+
+OpenCode Go is a subscription provider for a curated set of open coding models. Zoo Code loads the current model catalog from OpenCode Go and applies model-specific context, reasoning, caching, and pricing metadata.
+
+**Website:** [https://opencode.ai/docs/go/](https://opencode.ai/docs/go/)
+
+---
+
+## Getting an API Key
+
+1. Sign in to the [OpenCode console](https://opencode.ai/auth).
+2. Subscribe to OpenCode Go.
+3. Copy your API key.
+4. In Zoo Code settings, select **Opencode Go** and enter the API key.
+
+---
+
+## Available Models
+
+The catalog includes GLM-5.3 and Qwen3.8 Max:
+
+- **GLM-5.3**: 1M-token context window, 128K maximum output, and always-on reasoning with Low, High, and Max effort levels.
+- **Qwen3.8 Max**: Multimodal input, 1M-token context window, 128K maximum output, prompt caching, and preserved reasoning.
+
+The model list is fetched dynamically and may change. See the [OpenCode Go documentation](https://opencode.ai/docs/go/) for current availability, usage limits, and pricing.
+
+---
+
+## Configuration in Zoo Code
+
+1. Open Zoo Code settings using the gear icon (<Codicon name="gear" />).
+2. Select **Opencode Go** from the API Provider dropdown.
+3. Enter your OpenCode Go API key.
+4. Refresh the model list and select a model.
+
+Zoo Code automatically uses the API protocol required by the selected model.

From c6ea862ea8d5cc51b33e7024f1b0f7ae376a6fb9 Mon Sep 17 00:00:00 2001
From: "@taltas" <6816042+taltas@users.noreply.github.com>
Date: Fri, 14 Aug 2026 13:20:50 +0000
Subject: [PATCH 2/3] docs(provider): focus OpenCode Go guide on Qwen

---
 docs/providers/opencode-go.md | 8 ++------
 1 file changed, 2 insertions(+), 6 deletions(-)

diff --git a/docs/providers/opencode-go.md b/docs/providers/opencode-go.md
index 78347139..e6c0efd6 100644
--- a/docs/providers/opencode-go.md
+++ b/docs/providers/opencode-go.md
@@ -1,9 +1,8 @@
 ---
 sidebar_label: OpenCode Go
-description: Use OpenCode Go with Zoo Code for curated open coding models through a low-cost subscription.
+description: Use OpenCode Go with Zoo Code for curated open coding models, including Qwen3.8 Max, through a low-cost subscription.
 keywords:
   - opencode go
-  - glm-5.3
   - qwen3.8 max
   - coding models
   - zoo code
@@ -29,10 +28,7 @@ OpenCode Go is a subscription provider for a curated set of open coding models.
 
 ## Available Models
 
-The catalog includes GLM-5.3 and Qwen3.8 Max:
-
-- **GLM-5.3**: 1M-token context window, 128K maximum output, and always-on reasoning with Low, High, and Max effort levels.
-- **Qwen3.8 Max**: Multimodal input, 1M-token context window, 128K maximum output, prompt caching, and preserved reasoning.
+The catalog includes **Qwen3.8 Max**, with multimodal input, a 1M-token context window, 128K maximum output, prompt caching, and preserved reasoning.
 
 The model list is fetched dynamically and may change. See the [OpenCode Go documentation](https://opencode.ai/docs/go/) for current availability, usage limits, and pricing.
 

From 8ed22a316c96c124d9ef93a72981dd83ace8a468 Mon Sep 17 00:00:00 2001
From: "@taltas" <6816042+taltas@users.noreply.github.com>
Date: Fri, 14 Aug 2026 15:30:39 +0000
Subject: [PATCH 3/3] docs(provider): clarify Qwen reasoning continuity

---
 docs/providers/opencode-go.md | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)

diff --git a/docs/providers/opencode-go.md b/docs/providers/opencode-go.md
index e6c0efd6..c7dcbd20 100644
--- a/docs/providers/opencode-go.md
+++ b/docs/providers/opencode-go.md
@@ -28,7 +28,9 @@ OpenCode Go is a subscription provider for a curated set of open coding models.
 
 ## Available Models
 
-The catalog includes **Qwen3.8 Max**, with multimodal input, a 1M-token context window, 128K maximum output, prompt caching, and preserved reasoning.
+The catalog includes **Qwen3.8 Max**, with multimodal input, a 1M-token context window, 128K maximum output, prompt caching, and streamed reasoning output.
+
+Zoo Code displays the model's streamed reasoning output. OpenCode Go does not currently provide a usable signed-thinking continuity field, so that reasoning is not replayed across requests.
 
 The model list is fetched dynamically and may change. See the [OpenCode Go documentation](https://opencode.ai/docs/go/) for current availability, usage limits, and pricing.