[Feat] Add Kimi K3.1 Pro and Highspeed models

.changeset/kimi-k31-pro-models.md

@@ -0,0 +1,5 @@
+---
+'@roomote/web': patch
+---
+
+Add Kimi K3.1 Pro and Highspeed to recommended models on verified provider routes, including explicit Kimi for Coding runtime metadata.

apps/docs/models.mdx

@@ -252,6 +252,29 @@ version, and whether the error occurs on navigation or when opening a preset.
 Check migration logs for schema errors separately from invalid saved selections.
 Do not wipe model settings or reset the database to recover a preset.
 
+### Kimi K3.1 Pro
+
+**Kimi K3.1 Pro** is recommended through Moonshot AI, OpenRouter, Vercel AI
+Gateway, Kimi for Coding, and OpenCode Go. **Kimi K3.1 Pro Highspeed**
+is also available through Moonshot AI, Vercel AI Gateway, and Kimi
+for Coding.
+
+In **Settings > Models**, connect one of these providers and select the model.
+For manual configuration, use the provider's full task model ID:
+
+| Provider | Kimi K3.1 Pro | Kimi K3.1 Pro Highspeed |
+| --- | --- | --- |
+| Moonshot AI | `moonshotai/kimi-k3.1-pro` | `moonshotai/kimi-k3.1-pro-highspeed` |
+| OpenRouter | `openrouter/moonshotai/kimi-k3.1-pro` | — |
+| Vercel AI Gateway | `vercel/moonshotai/kimi-k3.1-pro` | `vercel/moonshotai/kimi-k3.1-pro-highspeed` |
+| Kimi for Coding | `kimi-for-coding/k3.1-pro` | `kimi-for-coding/k3.1-pro-highspeed` |
+| OpenCode Go | `opencode-go/kimi-k3.1-pro` | — |
+
+Kimi for Coding requires an eligible membership plan; access to the full 1M
+context window and Highspeed depends on the plan. Keep thinking enabled and
+start a new session when moving from an older Kimi model. See
+[Kimi's model and membership requirements](https://www.kimi.com/code/docs/en/kimi-code/models.html).
+
 ## Env-based setup
 
 Most deployments should configure providers from **Settings > Models**. Use

packages/types/src/kimi-for-coding-opencode-provider.ts

@@ -47,6 +47,28 @@ type KimiForCodingModelDefinition = {
 const KIMI_FOR_CODING_MODELS: Readonly<
   Record<string, KimiForCodingModelDefinition>
 > = {
+  'k3.1-pro': {
+    name: 'Kimi K3.1 Pro',
+    family: 'kimi-k3',
+    release_date: '2026-10-23',
+    attachment: true,
+    reasoning: true,
+    temperature: false,
+    tool_call: true,
+    modalities: { input: ['text', 'image', 'video'], output: ['text'] },
+    limit: { context: 1_048_576, output: 131_072 },
+  },
+  'k3.1-pro-highspeed': {
+    name: 'Kimi K3.1 Pro Highspeed',
+    family: 'kimi-k3',
+    release_date: '2026-10-23',
+    attachment: true,
+    reasoning: true,
+    temperature: false,
+    tool_call: true,
+    modalities: { input: ['text', 'image', 'video'], output: ['text'] },
+    limit: { context: 1_048_576, output: 131_072 },
+  },
   k3: {
     name: 'Kimi K3',
     family: 'kimi-k3',

packages/types/src/model-provider-config.ts

@@ -788,6 +788,8 @@ export const SETUP_MODEL_PROVIDER_CATALOG = [
       'deepseek-v4-pro-0813': 'vercel/deepseek/deepseek-v4-pro-0813',
       'glm-5-3-flash': 'vercel/zai/glm-5.3-flash',
       'glm-5-3': 'vercel/zai/glm-5.3',
+      'kimi-k3-1-pro': 'vercel/moonshotai/kimi-k3.1-pro',
+      'kimi-k3-1-pro-highspeed': 'vercel/moonshotai/kimi-k3.1-pro-highspeed',
       'kimi-k3': 'vercel/moonshotai/kimi-k3',
       'kimi-k2-7-code': 'vercel/moonshotai/kimi-k2.7-code',
       'qwen3-8-max': 'vercel/alibaba/qwen3.8-max',
@@ -1035,6 +1037,8 @@ export const SETUP_MODEL_PROVIDER_CATALOG = [
       linkLabel: 'Open Kimi API Platform keys',
     },
     suggestedTaskModels: mapRecommendedTaskModels({
+      'kimi-k3-1-pro': 'moonshotai/kimi-k3.1-pro',
+      'kimi-k3-1-pro-highspeed': 'moonshotai/kimi-k3.1-pro-highspeed',
       'kimi-k3': 'moonshotai/kimi-k3',
       'kimi-k2-7-code': 'moonshotai/kimi-k2.7-code',
     }),
@@ -1062,6 +1066,8 @@ export const SETUP_MODEL_PROVIDER_CATALOG = [
     // renamed upstream to `kimi-for-coding`, the rolling default coding model
     // (no longer K2.7, so it is not offered under that recommended name).
     suggestedTaskModels: mapRecommendedTaskModels({
+      'kimi-k3-1-pro': 'kimi-for-coding/k3.1-pro',
+      'kimi-k3-1-pro-highspeed': 'kimi-for-coding/k3.1-pro-highspeed',
       'kimi-k3': 'kimi-for-coding/k3',
     }),
     recommendedRoleModels: {
@@ -1143,6 +1149,7 @@ export const SETUP_MODEL_PROVIDER_CATALOG = [
       'glm-5-3': 'opencode-go/glm-5.3',
       'grok-4-7': 'opencode-go/grok-4.7',
       'kimi-k2-7-code': 'opencode-go/kimi-k2.7-code',
+      'kimi-k3-1-pro': 'opencode-go/kimi-k3.1-pro',
       'kimi-k3': 'opencode-go/kimi-k3',
       'minimax-m3': 'opencode-go/minimax-m3',
       'qwen3-8-max': 'opencode-go/qwen3.8-max',

packages/types/src/recommended-task-models.ts

@@ -69,6 +69,12 @@ export const RECOMMENDED_TASK_MODELS = [
   { id: 'glm-5-3-flash', displayName: 'GLM 5.3 Flash', family: 'GLM' },
   { id: 'glm-5-3', displayName: 'GLM 5.3', family: 'GLM' },
   { id: 'glm-5-2', displayName: 'GLM 5.2', family: 'GLM' },
+  { id: 'kimi-k3-1-pro', displayName: 'Kimi K3.1 Pro', family: 'Kimi' },
+  {
+    id: 'kimi-k3-1-pro-highspeed',
+    displayName: 'Kimi K3.1 Pro Highspeed',
+    family: 'Kimi',
+  },
   { id: 'kimi-k3', displayName: 'Kimi K3', family: 'Kimi' },
   { id: 'kimi-k2-7-code', displayName: 'Kimi K2.7 Code', family: 'Kimi' },
   { id: 'qwen3-8-max', displayName: 'Qwen3.8 Max', family: 'Qwen' },
@@ -129,6 +135,7 @@ export const OPENROUTER_RECOMMENDED_TASK_MODEL_SLUGS = {
   'deepseek-v4-pro-0813': 'openrouter/deepseek/deepseek-v4-pro-0813',
   'glm-5-3-flash': 'openrouter/z-ai/glm-5.3-flash',
   'glm-5-3': 'openrouter/z-ai/glm-5.3',
+  'kimi-k3-1-pro': 'openrouter/moonshotai/kimi-k3.1-pro',
   'kimi-k3': 'openrouter/moonshotai/kimi-k3',
   'kimi-k2-7-code': 'openrouter/moonshotai/kimi-k2.7-code',
   'qwen3-8-max': 'openrouter/qwen/qwen3.8-max-0902',