diff --git a/packages/opencode/src/provider/transform.ts b/packages/opencode/src/provider/transform.ts
index 0ebd8bbf59fe..e927af4ab26a 100644
--- a/packages/opencode/src/provider/transform.ts
+++ b/packages/opencode/src/provider/transform.ts
@@ -1000,7 +1000,11 @@ export function providerOptions(model: Provider.Model, options: { [x: string]: a
 }
 
 export function maxOutputTokens(model: Provider.Model): number {
-  return Math.min(model.limit.output, OUTPUT_TOKEN_MAX) || OUTPUT_TOKEN_MAX
+  const output = model.limit.output
+  if (!output || output <= 0) {
+    return OUTPUT_TOKEN_MAX
+  }
+  return Math.min(output, OUTPUT_TOKEN_MAX)
 }
 
 export function schema(model: Provider.Model, schema: JSONSchema.BaseSchema | JSONSchema7): JSONSchema7 {
diff --git a/packages/opencode/test/provider/transform.test.ts b/packages/opencode/test/provider/transform.test.ts
index e195d9b17720..2d83b01b93d6 100644
--- a/packages/opencode/test/provider/transform.test.ts
+++ b/packages/opencode/test/provider/transform.test.ts
@@ -3113,3 +3113,51 @@ describe("ProviderTransform.variants", () => {
     })
   })
 })
+
+describe("ProviderTransform.maxOutputTokens", () => {
+  const OUTPUT_TOKEN_MAX = 32000
+
+  const createMockModel = (overrides: any = {}) => ({
+    id: "test/model",
+    providerID: "test",
+    api: { id: "test", url: "https://test.com", npm: "@ai-sdk/test" },
+    name: "Test Model",
+    capabilities: { reasoning: false, toolcall: true, attachment: true },
+    limit: { context: 128000, output: 8192 },
+    status: "active",
+    options: {},
+    headers: {},
+    ...overrides,
+  })
+
+  test("should return model output limit when valid", () => {
+    const model = createMockModel({ limit: { context: 200000, output: 64000 } })
+    const result = ProviderTransform.maxOutputTokens(model)
+    // Should be min of model output limit (64000) and OUTPUT_TOKEN_MAX (32000)
+    expect(result).toBe(32000)
+  })
+
+  test("should return OUTPUT_TOKEN_MAX when model limit.output is undefined", () => {
+    const model = createMockModel({ limit: { context: 128000 } })
+    const result = ProviderTransform.maxOutputTokens(model)
+    expect(result).toBe(OUTPUT_TOKEN_MAX)
+  })
+
+  test("should return OUTPUT_TOKEN_MAX when model limit.output is 0", () => {
+    const model = createMockModel({ limit: { context: 128000, output: 0 } })
+    const result = ProviderTransform.maxOutputTokens(model)
+    expect(result).toBe(OUTPUT_TOKEN_MAX)
+  })
+
+  test("should return OUTPUT_TOKEN_MAX when model limit.output is negative", () => {
+    const model = createMockModel({ limit: { context: 128000, output: -1 } })
+    const result = ProviderTransform.maxOutputTokens(model)
+    expect(result).toBe(OUTPUT_TOKEN_MAX)
+  })
+
+  test("should return min of model limit and OUTPUT_TOKEN_MAX", () => {
+    const model = createMockModel({ limit: { context: 128000, output: 16000 } })
+    const result = ProviderTransform.maxOutputTokens(model)
+    expect(result).toBe(16000)
+  })
+})