chore(studio): default assistant to GPT-5.6 Luna (#49749)

## I have read the
[CONTRIBUTING.md](https://github.com/supabase/supabase/blob/master/CONTRIBUTING.md)
file.

YES

## What kind of change does this PR introduce?

Feature / chore: hide assistant model selection in the UI and default
chats to GPT-5.6 Luna.

## What is the current behavior?

The assistant composer exposes a model picker. Paid orgs default to
`gpt-5.3-codex`; everyone else defaults to `gpt-5.4-nano`.

## What is the new behavior?

- The model picker is hidden in the assistant composer and Explorer
home.
- Chats default to `gpt-5.6-luna` with `reasoningEffort: medium`.
- Model selection plumbing is kept (registry, entitlements, `setModel`,
generate-v4 request body) so a requested model can still be honored when
provided.
- Other completion endpoints still use `gpt-5.4-nano`.

## Additional context

Model selector UI can be re-enabled by passing `selectedModel` /
`onSelectModel` to `AssistantChatForm`.


<!-- This is an auto-generated comment: release notes by coderabbit.ai
-->

## Summary by CodeRabbit

* **New Features**
* Added support for the GPT-5.6 Luna model with medium reasoning
capability.
  * Made GPT-5.6 Luna the default assistant model.
* **Improvements**
* Simplified assistant chat by removing model selection from the primary
chat experience.
  * Updated model fallback behavior to use the standard assistant model.
* Chat forms can now optionally display model selection when configured.
* **Tests**
* Updated model coverage and assistant chat tests for the new defaults
and behavior.

<!-- end of auto-generated comment: release notes by coderabbit.ai -->

---------

Co-authored-by: Cursor <cursoragent@cursor.com>

apps/studio/components/interfaces/Explorer/ExplorerHomeTab.tsx

@@ -5,7 +5,6 @@ import { useCreateChat, useCreateNotebook, useCreateQuery } from './hooks'
 import { CHAT_TEMPLATES, NOTEBOOK_TEMPLATES } from './templates'
 import { ActionCard } from '@/components/layouts/Tabs/ActionCard'
 import { AssistantChatForm } from '@/components/ui/AIAssistantPanel/AssistantChatForm'
-import type { AssistantModel } from '@/state/ai-assistant-state'
 
 export const ExplorerHomeTab = () => {
   const { createNotebook } = useCreateNotebook()
@@ -13,7 +12,6 @@ export const ExplorerHomeTab = () => {
   const { createChat } = useCreateChat()
 
   const [value, setValue] = useState<string>('')
-  const [selectedModel, setSelectedModal] = useState<AssistantModel>('gpt-5.4-nano')
 
   return (
     <div className="bg-surface-100 h-full flex flex-col items-center justify-center">
@@ -31,9 +29,7 @@ export const ExplorerHomeTab = () => {
           placeholder="Ask anything about your project"
           value={value}
           onValueChange={(e) => setValue(e.target.value)}
-          selectedModel={selectedModel}
-          onSelectModel={setSelectedModal}
-          onSubmit={(message) => createChat({ initialMessage: message, model: selectedModel })}
+          onSubmit={(message) => createChat({ initialMessage: message })}
         />
 
         <section className="mt-6">

apps/studio/components/interfaces/Explorer/__tests__/hooks.test.tsx

@@ -9,7 +9,6 @@ const {
   mockPush,
   mockSelectChat,
   mockSetContext,
-  mockSetModel,
   mockWhenInitialized,
 } = vi.hoisted(() => ({
   mockCreateChat: vi.fn(() => 'chat-2'),
@@ -17,7 +16,6 @@ const {
   mockPush: vi.fn(),
   mockSelectChat: vi.fn(),
   mockSetContext: vi.fn(),
-  mockSetModel: vi.fn(),
   mockWhenInitialized: vi.fn(() => Promise.resolve()),
 }))
 
@@ -41,7 +39,6 @@ vi.mock('@/state/ai-assistant-state', () => ({
     createChat: mockCreateChat,
     selectChat: mockSelectChat,
     setContext: mockSetContext,
-    setModel: mockSetModel,
   }),
   whenAiAssistantInitialized: () => mockWhenInitialized(),
 }))
@@ -59,7 +56,6 @@ describe('useCreateChat', () => {
       await result.current.createChat({
         name: 'Investigate errors',
         initialMessage: 'What happened?',
-        model: 'gpt-5.4-nano',
       })
     })
 
@@ -72,7 +68,6 @@ describe('useCreateChat', () => {
       name: 'Investigate errors',
       initialMessage: 'What happened?',
     })
-    expect(mockSetModel).toHaveBeenCalledWith('gpt-5.4-nano')
     expect(mockPush).toHaveBeenCalledWith('/project/default/explorer/chat/chat-2')
     expect(mockSelectChat).not.toHaveBeenCalled()
 
@@ -82,7 +77,7 @@ describe('useCreateChat', () => {
     expect(mockSelectChat).not.toHaveBeenCalled()
   })
 
-  // Hydration replaces the chat map and the model wholesale, so a chat created mid-load would be
+  // Hydration replaces the chat map wholesale, so a chat created mid-load would be
   // dropped the moment the persisted state lands
   it('waits for the assistant state to hydrate before creating the chat', async () => {
     let resolveHydration = () => {}
@@ -97,11 +92,10 @@ describe('useCreateChat', () => {
 
     let created: Promise<string | undefined> | undefined
     await act(async () => {
-      created = result.current.createChat({ name: 'Investigate errors', model: 'gpt-5.4-nano' })
+      created = result.current.createChat({ name: 'Investigate errors' })
     })
 
     expect(mockCreateChat).not.toHaveBeenCalled()
-    expect(mockSetModel).not.toHaveBeenCalled()
     expect(mockPush).not.toHaveBeenCalled()
 
     await act(async () => {
@@ -113,7 +107,6 @@ describe('useCreateChat', () => {
       name: 'Investigate errors',
       initialMessage: undefined,
     })
-    expect(mockSetModel).toHaveBeenCalledWith('gpt-5.4-nano')
     expect(mockPush).toHaveBeenCalledWith('/project/default/explorer/chat/chat-2')
   })
 })

apps/studio/components/ui/AIAssistantPanel/AssistantChat.tsx

@@ -32,23 +32,16 @@ import { Markdown } from '@/components/interfaces/Markdown'
 import { useCheckOpenAIKeyQuery } from '@/data/ai/check-api-key-query'
 import { useRateMessageMutation } from '@/data/ai/rate-message-mutation'
 import { useTablesQuery } from '@/data/tables/tables-query'
-import { useCheckEntitlements } from '@/hooks/misc/useCheckEntitlements'
 import { useLocalStorageQuery } from '@/hooks/misc/useLocalStorage'
 import { useOrgAiOptInLevel } from '@/hooks/misc/useOrgOptedIntoAi'
 import { useSelectedOrganizationQuery } from '@/hooks/misc/useSelectedOrganization'
 import { useSelectedProjectQuery } from '@/hooks/misc/useSelectedProject'
 import type { AssistantMessageMetadata } from '@/lib/ai/assistant-message-metadata'
 import { getParallelApprovalIdsToReject } from '@/lib/ai/message-utils'
-import {
-  DEFAULT_ASSISTANT_BASE_MODEL_ID,
-  defaultAssistantModelId,
-  isAssistantBaseModelId,
-  isKnownAssistantModelId,
-} from '@/lib/ai/model.utils'
 import { IS_PLATFORM } from '@/lib/constants'
 import { uuidv4 } from '@/lib/helpers'
 import { useTrack } from '@/lib/telemetry/track'
-import type { AssistantModel, SqlSnippet } from '@/state/ai-assistant-state'
+import type { SqlSnippet } from '@/state/ai-assistant-state'
 import { useAiAssistantState, useAiAssistantStateSnapshot } from '@/state/ai-assistant-state'
 import { SHORTCUT_IDS } from '@/state/shortcuts/registry'
 import { useShortcut } from '@/state/shortcuts/useShortcut'
@@ -109,26 +102,6 @@ export const AssistantChat = ({
     enabled: shortcutsEnabled,
   })
 
-  const { hasAccess: hasAccessToAdvanceModel, isLoading: isLoadingEntitlements } =
-    useCheckEntitlements('assistant.advance_model')
-
-  const selectedModel = useMemo<AssistantModel>(() => {
-    // While entitlements are loading, use the stored model without enforcing access
-    if (isLoadingEntitlements) {
-      return snap.model ?? DEFAULT_ASSISTANT_BASE_MODEL_ID
-    }
-
-    const defaultModel = defaultAssistantModelId(hasAccessToAdvanceModel)
-    const model = snap.model ?? defaultModel
-
-    if (!isKnownAssistantModelId(model)) return defaultModel
-    if (!hasAccessToAdvanceModel && !isAssistantBaseModelId(model)) {
-      return DEFAULT_ASSISTANT_BASE_MODEL_ID
-    }
-
-    return model
-  }, [isLoadingEntitlements, hasAccessToAdvanceModel, snap.model])
-
   const [updatedOptInSinceMCP] = useLocalStorageQuery(
     LOCAL_STORAGE_KEYS.AI_ASSISTANT_MCP_OPT_IN,
     false
@@ -652,7 +625,6 @@ export const AssistantChat = ({
                 </div>
               </div>
             )}
-
             {disablePrompts && (
               <Admonition
                 showIcon={false}
@@ -706,8 +678,6 @@ export const AssistantChat = ({
                 composerContext?.onSetSqlSnippets?.(newSnippets)
               }}
               includeSnippetsInMessage={includeSnippetsInMessage}
-              selectedModel={selectedModel}
-              onSelectModel={(model) => snap.setModel(model)}
             />
           </div>
         </div>

apps/studio/components/ui/AIAssistantPanel/AssistantChatForm.tsx

@@ -46,9 +46,9 @@ export interface FormProps {
   /* If currently editing an existing message */
   isEditing?: boolean
   /* The currently selected AI model */
-  selectedModel: AssistantModelId
+  selectedModel?: AssistantModelId
   /* Callback when a model is chosen */
-  onSelectModel: (model: AssistantModelId) => void
+  onSelectModel?: (model: AssistantModelId) => void
 }
 
 const AssistantChatFormComponent = forwardRef<HTMLFormElement, FormProps>(
@@ -99,6 +99,7 @@ const AssistantChatFormComponent = forwardRef<HTMLFormElement, FormProps>(
     }
 
     const canSubmit = !disabled && !loading && !!value
+    const showModelSelector = selectedModel !== undefined && onSelectModel !== undefined
 
     return (
       <div className="w-full">
@@ -131,10 +132,17 @@ const AssistantChatFormComponent = forwardRef<HTMLFormElement, FormProps>(
             onChange={(event) => onValueChange(event)}
             onKeyDown={handleKeyDown}
           />
-          <div className="absolute inset-x-1.5 bottom-1.5 flex items-center justify-between pointer-events-none">
-            <div className="pointer-events-auto">
-              <ModelSelector selectedModel={selectedModel} onSelectModel={onSelectModel} />
-            </div>
+          <div
+            className={cn(
+              'absolute inset-x-1.5 bottom-1.5 flex items-center pointer-events-none',
+              showModelSelector ? 'justify-between' : 'justify-end'
+            )}
+          >
+            {showModelSelector && (
+              <div className="pointer-events-auto">
+                <ModelSelector selectedModel={selectedModel} onSelectModel={onSelectModel} />
+              </div>
+            )}
 
             <div className="flex gap-3 items-center pointer-events-auto">
               {loading ? (

apps/studio/lib/ai/model.test.ts

@@ -70,9 +70,8 @@ describe('getModel', () => {
     expect(error).toEqual(new Error('OPENAI_API_KEY not available'))
   })
 
-  it('returns openai gpt-5.3-codex when hasAccessToAdvanceModel and not throttled', async () => {
+  it('returns openai gpt-5.3-codex with reasoning effort', async () => {
     vi.stubEnv('OPENAI_API_KEY', 'test-key')
-    vi.stubEnv('IS_THROTTLED', 'false')
 
     const { modelParams, error } = await getModel({
       provider: 'openai',

apps/studio/lib/ai/model.utils.test.ts

@@ -47,8 +47,9 @@ describe('model.utils', () => {
     it('should have openai provider with models', () => {
       expect(PROVIDERS.openai).toBeDefined()
       expect(PROVIDERS.openai.models).toBeDefined()
-      expect(Object.keys(PROVIDERS.openai.models)).toContain('gpt-5.3-codex')
+      expect(Object.keys(PROVIDERS.openai.models)).toContain('gpt-5.6-luna')
       expect(Object.keys(PROVIDERS.openai.models)).toContain('gpt-5.4-nano')
+      expect(Object.keys(PROVIDERS.openai.models)).toContain('gpt-5.3-codex')
     })
 
     it('should have exactly one default model per provider', () => {
@@ -111,20 +112,23 @@ describe('model.utils', () => {
     })
 
     it('defaults should satisfy unions', () => {
-      expect(DEFAULT_ASSISTANT_BASE_MODEL_ID).toBe('gpt-5.4-nano')
+      expect(DEFAULT_ASSISTANT_BASE_MODEL_ID).toBe('gpt-5.6-luna')
       expect(DEFAULT_ASSISTANT_ADVANCE_MODEL_ID).toBe('gpt-5.3-codex')
       expect(defaultAssistantModelId(false)).toBe(DEFAULT_ASSISTANT_BASE_MODEL_ID)
-      expect(defaultAssistantModelId(true)).toBe(DEFAULT_ASSISTANT_ADVANCE_MODEL_ID)
+      expect(defaultAssistantModelId(true)).toBe(DEFAULT_ASSISTANT_BASE_MODEL_ID)
     })
 
     it('isAssistantBaseModelId / isAdvanceOnlyModelId', () => {
+      expect(isAssistantBaseModelId('gpt-5.6-luna')).toBe(true)
       expect(isAssistantBaseModelId('gpt-5.4-nano')).toBe(true)
       expect(isAssistantBaseModelId('gpt-5.3-codex')).toBe(false)
       expect(isAdvanceOnlyModelId('gpt-5.3-codex')).toBe(true)
       expect(isAdvanceOnlyModelId('gpt-5.4-nano')).toBe(false)
+      expect(isAdvanceOnlyModelId('gpt-5.6-luna')).toBe(false)
     })
 
     it('isKnownAssistantModelId', () => {
+      expect(isKnownAssistantModelId('gpt-5.6-luna')).toBe(true)
       expect(isKnownAssistantModelId('gpt-5.4-nano')).toBe(true)
       expect(isKnownAssistantModelId('gpt-5.3-codex')).toBe(true)
       expect(isKnownAssistantModelId('gpt-5')).toBe(false)
@@ -133,28 +137,27 @@ describe('model.utils', () => {
     })
 
     it('getAssistantModelEntry returns config for known ids', () => {
+      expect(getAssistantModelEntry('gpt-5.6-luna').reasoningEffort).toBe('medium')
       expect(getAssistantModelEntry('gpt-5.4-nano').reasoningEffort).toBe('low')
       expect(getAssistantModelEntry('gpt-5.3-codex').reasoningEffort).toBe('low')
-      expect(getAssistantModelEntry('gpt-5.4-nano')).toEqual(
-        ASSISTANT_MODELS.find((m) => m.id === 'gpt-5.4-nano')
+      expect(getAssistantModelEntry('gpt-5.6-luna')).toEqual(
+        ASSISTANT_MODELS.find((m) => m.id === 'gpt-5.6-luna')
       )
     })
 
     it('DEFAULT_COMPLETION_MODEL is gpt-5.4-nano with no reasoning effort', () => {
-      expect(DEFAULT_COMPLETION_MODEL.id).toBe(DEFAULT_ASSISTANT_BASE_MODEL_ID)
+      expect(DEFAULT_COMPLETION_MODEL.id).toBe('gpt-5.4-nano')
       expect(DEFAULT_COMPLETION_MODEL.reasoningEffort).toBe('none')
     })
 
     it('openaiModelEntry enforces valid reasoning effort at compile time', () => {
-      // Valid: supported effort level
       const withEffort = openaiModelEntry({
-        id: 'gpt-5.4-nano',
+        id: 'gpt-5.6-luna',
         reasoningEffort: 'low',
       })
       expect(withEffort.reasoningEffort).toBe('low')
 
-      // Valid: no effort
-      const withoutEffort = openaiModelEntry({ id: 'gpt-5.4-nano' })
+      const withoutEffort = openaiModelEntry({ id: 'gpt-5.6-luna' })
       expect(withoutEffort.reasoningEffort).toBeUndefined()
     })
   })

apps/studio/lib/ai/model.utils.ts

@@ -2,17 +2,19 @@ export type ProviderName = 'bedrock' | 'openai'
 
 export type BedrockModel = 'anthropic.claude-3-7-sonnet-20250219-v1:0' | 'openai.gpt-oss-120b-1:0'
 
-export type OpenAIModelId = 'gpt-5.4-nano' | 'gpt-5.3-codex'
+export type OpenAIModelId = 'gpt-5.4-nano' | 'gpt-5.3-codex' | 'gpt-5.6-luna'
 
 // Source: https://developers.openai.com/api/docs/guides/reasoning + per-model pages
-export type ReasoningEffort = 'none' | 'minimal' | 'low' | 'medium' | 'high' | 'xhigh'
+export type ReasoningEffort = 'none' | 'minimal' | 'low' | 'medium' | 'high' | 'xhigh' | 'max'
 
 // Per-model reasoning effort compatibility.
 // Sources: https://developers.openai.com/api/docs/models/gpt-5.4-nano
 //          https://developers.openai.com/api/docs/models/gpt-5.3-codex
+//          https://developers.openai.com/api/docs/models/gpt-5.6-luna
 type ModelReasoningSupport = {
   'gpt-5.4-nano': 'none' | 'low' | 'medium' | 'high' | 'xhigh'
   'gpt-5.3-codex': 'low' | 'medium' | 'high' | 'xhigh'
+  'gpt-5.6-luna': 'none' | 'low' | 'medium' | 'high' | 'xhigh' | 'max'
 }
 
 type ReasoningEffortFor<ModelId extends OpenAIModelId> = ModelId extends keyof ModelReasoningSupport
@@ -59,6 +61,11 @@ export const LOGS_REWRITE_MODEL = openaiModelEntry({
 // Single source of truth for all Assistant chat model variants and their reasoning levels.
 // Models with requiresAdvanceModelEntitlement false are available to all users; true requires the assistant.advance_model entitlement.
 export const ASSISTANT_MODELS = [
+  openaiModelEntry({
+    id: 'gpt-5.6-luna',
+    requiresAdvanceModelEntitlement: false,
+    reasoningEffort: 'medium',
+  }),
   openaiModelEntry({
     id: 'gpt-5.4-nano',
     requiresAdvanceModelEntitlement: false,
@@ -82,14 +89,12 @@ const ASSISTANT_MODELS_MAP = Object.fromEntries(ASSISTANT_MODELS.map((m) => [m.i
   (typeof ASSISTANT_MODELS)[number]
 >
 
-export const DEFAULT_ASSISTANT_BASE_MODEL_ID = 'gpt-5.4-nano' satisfies AssistantBaseModelId
+export const DEFAULT_ASSISTANT_BASE_MODEL_ID = 'gpt-5.6-luna' satisfies AssistantBaseModelId
 
 export const DEFAULT_ASSISTANT_ADVANCE_MODEL_ID = 'gpt-5.3-codex' satisfies AssistantModelId
 
-export function defaultAssistantModelId(hasAccessToAdvanceModel: boolean): AssistantModelId {
-  return hasAccessToAdvanceModel
-    ? DEFAULT_ASSISTANT_ADVANCE_MODEL_ID
-    : DEFAULT_ASSISTANT_BASE_MODEL_ID
+export function defaultAssistantModelId(_hasAccessToAdvanceModel: boolean): AssistantModelId {
+  return DEFAULT_ASSISTANT_BASE_MODEL_ID
 }
 
 export function isKnownAssistantModelId(id: string): id is AssistantModelId {
@@ -153,6 +158,7 @@ export const PROVIDERS: ProviderRegistry = {
   },
   openai: {
     models: {
+      'gpt-5.6-luna': { default: false },
       'gpt-5.3-codex': { default: false },
       'gpt-5.4-nano': { default: true },
     },

apps/studio/pages/api/ai/sql/generate-v4.ts

@@ -18,7 +18,6 @@ import { generateAssistantResponse } from '@/lib/ai/generate-assistant-response'
 import { isExplorerEnabled } from '@/lib/ai/is-explorer-enabled'
 import { getModel } from '@/lib/ai/model'
 import {
-  DEFAULT_ASSISTANT_ADVANCE_MODEL_ID,
   DEFAULT_ASSISTANT_BASE_MODEL_ID,
   getAssistantModelEntry,
   isAssistantBaseModelId,
@@ -157,7 +156,7 @@ async function handlePost(req: NextApiRequest, res: NextApiResponse, claims?: Jw
 
   const envThrottled = process.env.IS_THROTTLED !== 'false'
 
-  let effectiveModel: AssistantModelId = requestedModel ?? DEFAULT_ASSISTANT_ADVANCE_MODEL_ID
+  let effectiveModel: AssistantModelId = requestedModel ?? DEFAULT_ASSISTANT_BASE_MODEL_ID
   if (!hasAccessToAdvanceModel || (envThrottled && !isAssistantBaseModelId(effectiveModel))) {
     effectiveModel = DEFAULT_ASSISTANT_BASE_MODEL_ID
   }

apps/studio/state/ai-assistant-state.hooks.test.tsx

@@ -53,7 +53,6 @@ describe('useAiAssistantChatList', () => {
       state.loadPersistedState({
         projectRef: 'default',
         activeChatId: 'persisted-chat',
-        model: state.model,
         chats: {
           'persisted-chat': {
             id: 'persisted-chat',

apps/studio/state/ai-assistant-state.test.ts

@@ -146,7 +146,6 @@ describe('whenAiAssistantInitialized', () => {
     state.loadPersistedState({
       projectRef: 'default',
       activeChatId: 'persisted-chat',
-      model: state.model,
       chats: {
         'persisted-chat': {
           id: 'persisted-chat',