feat(providers): refresh curated model catalogs and retire Imagen 4

README.md

@@ -557,7 +557,7 @@ the current catalog at **[radar.omniroute.online/planos](https://radar.omniroute
 - **🧠 Memory you control** — off by default, opt-in int8 vector quantization + typed decay, per-request `x-omniroute-no-memory`. → [Memory](docs/frameworks/MEMORY.md)
 - **🛡️ Security** — prompt-injection guard on every LLM route (red-team suite), opt-in credential-masking guardrail (redacts leaked API keys/secrets in both directions), free DuckDuckGo last-resort web search, and an optional OIDC login gate for the dashboard (password login always stays available). → [Guardrails](docs/security/GUARDRAILS.md)
 - **🖼️ New endpoints** — `/v1/ocr` (Mistral OCR) and `/v1/audio/translations` (Whisper-style) round out the media surface. → [API Reference](docs/reference/API_REFERENCE.md)
-- **🎨 Image / video / audio generation** — one API for media: xAI Grok Imagine & Novita AI video, ComfyUI, Freepik, Adobe Firefly, Microsoft Designer, Google Imagen, Segmind, EdgeTTS. → [API Reference](docs/reference/API_REFERENCE.md)
+- **🎨 Image / video / audio generation** — one API for media: xAI Grok Imagine & Novita AI video, ComfyUI, Freepik, Adobe Firefly, Microsoft Designer, Segmind, EdgeTTS. → [API Reference](docs/reference/API_REFERENCE.md)
 - **🌍 Deployment & ops** — reverse-proxy `basePath`, browser-language auto-detect, per-key device tracking, root-less MITM trust, zh-TW localization. → [Environment](docs/reference/ENVIRONMENT.md)
 - **🤝 More providers & agents** — Cursor Cloud Agent, Grok Build (xAI) with browser + OAuth login, Ollama first-class card, Claude Opus 5 & Sonnet 5, Kimi official partnership (Code/Web/Moonshot), Zed, Requesty, SenseNova, Yuanbao, Agnes AI… and a refreshed **340-provider catalog**. → [Providers](docs/reference/PROVIDER_REFERENCE.md)
 - **📡 Routing transparency** — every response carries an `X-OmniRoute-Decision` header naming the strategy/provider/latency that served it, a new `cache-optimized` combo strategy + Auto-Combo `cacheAffinity` factor route repeat requests back to the connection holding the cached prefix, and a read-only `/v1/auto-combo/{channel}/candidates` endpoint exposes an `auto/*` channel's live candidate pool. → [Auto-Combo](docs/routing/AUTO-COMBO.md)

open-sse/config/agyModels.ts

@@ -2,8 +2,8 @@
 //
 // These models are pinned from the live `:fetchAvailableModels` endpoint
 // (https://daily-cloudcode-pa.googleapis.com/v1internal:fetchAvailableModels) using a
-// real `agy` consumer-OAuth token. The public catalog exposes the upstream Gemini 3.6
-// and 3.5 Flash ids verbatim; the shared Antigravity executor dispatches them unchanged.
+// real `agy` consumer-OAuth token. The public catalog exposes the upstream Gemini 3.7
+// Flash ids verbatim; the shared Antigravity executor dispatches them unchanged.
 //
 // The `agy` provider reuses the `antigravity` executor/translator (identical backend),
 // but keeps its own catalog so the CLI and IDE model surfaces can evolve independently.
@@ -12,48 +12,29 @@
 // they are not chat-callable.
 
 export const AGY_PUBLIC_MODELS = Object.freeze([
-  // Gemini 3.6 Flash tiers. The live endpoint selects High by default and advertises
-  // all three ids to both the IDE 2.1.1 and CLI 1.1.x clients.
+  // Gemini 3.7 Flash tiers. The live endpoint selects High by default and advertises
+  // all three ids to both the IDE 2.5.5 and CLI 1.1.x clients.
   {
-    id: "gemini-3.6-flash-high",
-    name: "Gemini 3.6 Flash (High)",
+    id: "gemini-3.7-flash-high",
+    name: "Gemini 3.7 Flash (High)",
     contextLength: 1048576,
     maxOutputTokens: 65536,
     supportsReasoning: true,
     supportsVision: true,
     toolCalling: true,
   },
   {
-    id: "gemini-3.6-flash-medium",
-    name: "Gemini 3.6 Flash (Medium)",
+    id: "gemini-3.7-flash-medium",
+    name: "Gemini 3.7 Flash (Medium)",
     contextLength: 1048576,
     maxOutputTokens: 65536,
     supportsReasoning: true,
     supportsVision: true,
     toolCalling: true,
   },
   {
-    id: "gemini-3.6-flash-low",
-    name: "Gemini 3.6 Flash (Low)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  // Claude (Antigravity backend).
-  {
-    id: "claude-opus-4-6-thinking",
-    name: "Claude Opus 4.6 (Thinking)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "claude-sonnet-4-6",
-    name: "Claude Sonnet 4.6 (Thinking)",
+    id: "gemini-3.7-flash-low",
+    name: "Gemini 3.7 Flash (Low)",
     contextLength: 1048576,
     maxOutputTokens: 65536,
     supportsReasoning: true,
@@ -80,74 +61,31 @@ export const AGY_PUBLIC_MODELS = Object.freeze([
     toolCalling: true,
   },
   {
-    id: "gemini-3-flash-agent",
-    name: "Gemini 3.5 Flash (High)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-3.5-flash-low",
-    name: "Gemini 3.5 Flash (Medium)",
+    id: "gemini-3.1-flash-lite",
+    name: "Gemini 3.1 Flash Lite",
     contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
+    maxOutputTokens: 65535,
     toolCalling: true,
   },
+  // Claude (Antigravity backend).
   {
-    id: "gemini-3.5-flash-extra-low",
-    name: "Gemini 3.5 Flash (Low)",
+    id: "claude-opus-4-6-thinking",
+    name: "Claude Opus 4.6 (Thinking)",
     contextLength: 1048576,
     maxOutputTokens: 65536,
     supportsReasoning: true,
     supportsVision: true,
     toolCalling: true,
   },
-  // Gemini 3.7 Flash: single callable public model (upstream exposes only
-  // gemini-3.7-flash-tiered; suffixed tier ids 404). One entry so it does not
-  // collide under the #3696 public-id uniqueness invariant.
   {
-    id: "gemini-3.7-flash",
-    name: "Gemini 3.7 Flash",
+    id: "claude-sonnet-4-6",
+    name: "Claude Sonnet 4.6 (Thinking)",
     contextLength: 1048576,
     maxOutputTokens: 65536,
     supportsReasoning: true,
     supportsVision: true,
     toolCalling: true,
   },
-  {
-    id: "gemini-3.1-flash-lite",
-    name: "Gemini 3.1 Flash Lite",
-    contextLength: 1048576,
-    maxOutputTokens: 65535,
-    toolCalling: true,
-  },
-  // Gemini 2.5
-  {
-    id: "gemini-2.5-flash-thinking",
-    name: "Gemini 2.5 Flash Thinking",
-    contextLength: 1048576,
-    maxOutputTokens: 65535,
-    supportsReasoning: true,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-2.5-flash",
-    name: "Gemini 2.5 Flash",
-    contextLength: 1048576,
-    maxOutputTokens: 65535,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-2.5-flash-lite",
-    name: "Gemini 2.5 Flash Lite",
-    contextLength: 1048576,
-    maxOutputTokens: 65535,
-    toolCalling: true,
-  },
   // GPT-OSS
   {
     id: "gpt-oss-120b-medium",
@@ -161,6 +99,21 @@ export const AGY_PUBLIC_MODELS = Object.freeze([
 
 const AGY_PUBLIC_MODEL_IDS = new Set(AGY_PUBLIC_MODELS.map((model) => model.id));
 const AGY_NON_CHAT_MODEL_IDS = new Set(["tab_flash_lite_preview", "tab_jump_flash_lite_preview"]);
+const AGY_RETIRED_MODEL_IDS = new Set([
+  "gemini-3.6-flash-high",
+  "gemini-3.6-flash-medium",
+  "gemini-3.6-flash-low",
+  "gemini-3-flash-agent",
+  "gemini-3.5-flash-extra-low",
+  "gemini-3.5-flash-low",
+  "gemini-3.5-flash-high",
+  "gemini-3.5-flash-medium",
+  "gemini-3.5-flash-preview",
+  "gemini-2.5-pro",
+  "gemini-2.5-flash-thinking",
+  "gemini-2.5-flash",
+  "gemini-2.5-flash-lite",
+]);
 
 const AGY_CLIENT_VISIBLE_MODEL_NAMES = Object.freeze(
   AGY_PUBLIC_MODELS.reduce<Record<string, string>>((acc, model) => {
@@ -178,5 +131,5 @@ export function isUserCallableAgyModelId(modelId: string): boolean {
 }
 
 export function isDiscoverableAgyModelId(modelId: string): boolean {
-  return !!modelId && !AGY_NON_CHAT_MODEL_IDS.has(modelId);
+  return !!modelId && !AGY_NON_CHAT_MODEL_IDS.has(modelId) && !AGY_RETIRED_MODEL_IDS.has(modelId);
 }

open-sse/config/antigravityModelAliases.ts

@@ -1,7 +1,7 @@
 export const ANTIGRAVITY_PUBLIC_MODELS = Object.freeze([
-  // Gemini 3.7 Flash tiers listed by the current official Antigravity model catalog
-  // alongside the existing Gemini 3.6 tiers. Keep the upstream model ids unchanged so
-  // discovery and execution address the same models selected by the native client.
+  // Gemini 3.7 Flash tiers listed by the current official Antigravity model catalog.
+  // Keep the upstream model ids unchanged so discovery and execution address the same
+  // models selected by the native client.
   {
     id: "gemini-3.7-flash-high",
     name: "Gemini 3.7 Flash (High)",
@@ -20,51 +20,9 @@ export const ANTIGRAVITY_PUBLIC_MODELS = Object.freeze([
     supportsVision: true,
     toolCalling: true,
   },
-  // Gemini 3.6 Flash tiers retained alongside the newer Gemini 3.7 tiers.
   {
-    id: "gemini-3.6-flash-high",
-    name: "Gemini 3.6 Flash (High)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-3.6-flash-medium",
-    name: "Gemini 3.6 Flash (Medium)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-3.6-flash-low",
-    name: "Gemini 3.6 Flash (Low)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  // Claude (Antigravity backend). The `agy` provider already ships these from the live
-  // :fetchAvailableModels probe (see agyModels.ts) and discussion #3184 confirmed they
-  // are user-callable through the `antigravity` OAuth provider too — same backend.
-  // `antigravity/claude-opus-4-6-thinking` and `antigravity/claude-sonnet-4-6` both work.
-  // They are upstream IDs, so no alias remapping is required.
-  {
-    id: "claude-opus-4-6-thinking",
-    name: "Claude Opus 4.6 (Thinking)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
-  {
-    id: "claude-sonnet-4-6",
-    name: "Claude Sonnet 4.6 (Thinking)",
+    id: "gemini-3.7-flash-low",
+    name: "Gemini 3.7 Flash (Low)",
     contextLength: 1048576,
     maxOutputTokens: 65536,
     supportsReasoning: true,
@@ -92,78 +50,36 @@ export const ANTIGRAVITY_PUBLIC_MODELS = Object.freeze([
     supportsVision: true,
     toolCalling: true,
   },
-  // Gemini 3.5 Flash tiers exposed by Antigravity's model selector. Public ids match
-  // fetchAvailableModels and are forwarded upstream unchanged:
-  //   High   -> gemini-3-flash-agent       (displayName: Gemini 3.5 Flash (High))
-  //   Medium -> gemini-3.5-flash-low       (displayName: Gemini 3.5 Flash (Medium))
-  //   Low    -> gemini-3.5-flash-extra-low (displayName: Gemini 3.5 Flash (Low))
-  {
-    id: "gemini-3-flash-agent",
-    name: "Gemini 3.5 Flash (High)",
-    contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
-    toolCalling: true,
-  },
   {
-    id: "gemini-3.5-flash-low",
-    name: "Gemini 3.5 Flash (Medium)",
+    id: "gemini-3.1-flash-lite",
+    name: "Gemini 3.1 Flash Lite",
     contextLength: 1048576,
-    maxOutputTokens: 65536,
-    supportsReasoning: true,
-    supportsVision: true,
+    maxOutputTokens: 65535,
     toolCalling: true,
   },
+  // Claude (Antigravity backend). The `agy` provider already ships these from the live
+  // :fetchAvailableModels probe (see agyModels.ts) and discussion #3184 confirmed they
+  // are user-callable through the `antigravity` OAuth provider too — same backend.
+  // `antigravity/claude-opus-4-6-thinking` and `antigravity/claude-sonnet-4-6` both work.
+  // They are upstream IDs, so no alias remapping is required.
   {
-    id: "gemini-3.5-flash-extra-low",
-    name: "Gemini 3.5 Flash (Low)",
+    id: "claude-opus-4-6-thinking",
+    name: "Claude Opus 4.6 (Thinking)",
     contextLength: 1048576,
     maxOutputTokens: 65536,
     supportsReasoning: true,
     supportsVision: true,
     toolCalling: true,
   },
-  // Gemini 3.7 Flash: Antigravity's live catalog exposes a single upstream id
-  // gemini-3.7-flash-tiered; the suffixed tier ids 404 upstream. Kept as one
-  // callable public model so it does not collide with the #3696 uniqueness invariant.
   {
-    id: "gemini-3.7-flash",
-    name: "Gemini 3.7 Flash",
+    id: "claude-sonnet-4-6",
+    name: "Claude Sonnet 4.6 (Thinking)",
     contextLength: 1048576,
     maxOutputTokens: 65536,
     supportsReasoning: true,
     supportsVision: true,
     toolCalling: true,
   },
-  {
-    id: "gemini-3.1-flash-lite",
-    name: "Gemini 3.1 Flash Lite",
-    contextLength: 1048576,
-    maxOutputTokens: 65535,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-2.5-flash-thinking",
-    name: "Gemini 2.5 Flash Thinking",
-    contextLength: 1048576,
-    maxOutputTokens: 65535,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-2.5-flash",
-    name: "Gemini 2.5 Flash",
-    contextLength: 1048576,
-    maxOutputTokens: 65535,
-    toolCalling: true,
-  },
-  {
-    id: "gemini-2.5-flash-lite",
-    name: "Gemini 2.5 Flash Lite",
-    contextLength: 1048576,
-    maxOutputTokens: 65535,
-    toolCalling: true,
-  },
   {
     id: "gpt-oss-120b-medium",
     name: "GPT-OSS 120B (Medium)",
@@ -175,12 +91,6 @@ export const ANTIGRAVITY_PUBLIC_MODELS = Object.freeze([
 ]);
 
 export const ANTIGRAVITY_MODEL_ALIASES = Object.freeze({
-  // Gemini 3.7 Flash: the live catalog (fetchAvailableModels on daily-cloudcode-pa)
-  // exposes a single upstream id `gemini-3.7-flash-tiered`; the agy CLI maps all
-  // display tiers (high/medium/low) to it. Verified 200 OK with thinking_level and
-  // thinkingBudget configs. The suffixed ids 404 upstream ("Requested entity was not found").
-  // Exposed as ONE callable model (see #3696: public ids must be unique upstream ids).
-  "gemini-3.7-flash": "gemini-3.7-flash-tiered",
   // gemini-3.1-pro-low is not aliased: the upstream accepts it verbatim.
   // gemini-3.1-pro-high: the discovery slot returns HTTP 400 on v1internal;
   // the live upstream id is gemini-pro-agent (see ANTIGRAVITY_PUBLIC_MODELS).
@@ -249,10 +159,19 @@ const ANTIGRAVITY_NON_CHAT_MODEL_IDS = new Set([
 const ANTIGRAVITY_RETIRED_MODEL_IDS = new Set([
   "gemini-3-pro-preview",
   "gemini-3.1-pro",
+  "gemini-3.6-flash-high",
+  "gemini-3.6-flash-medium",
+  "gemini-3.6-flash-low",
+  "gemini-3-flash-agent",
+  "gemini-3.5-flash-extra-low",
+  "gemini-3.5-flash-low",
   "gemini-3.5-flash-high",
   "gemini-3.5-flash-medium",
   "gemini-3.5-flash-preview",
   "gemini-2.5-pro",
+  "gemini-2.5-flash-thinking",
+  "gemini-2.5-flash",
+  "gemini-2.5-flash-lite",
   "gemini-2.5-computer-use-preview-10-2025",
 ]);
 
@@ -281,7 +200,12 @@ const ANTIGRAVITY_DROPPED_QUOTA_BUCKETS = new Set<string>([
  */
 export function toClientAntigravityQuotaModelId(modelId: string): string | null {
   if (!modelId) return null;
-  if (ANTIGRAVITY_DROPPED_QUOTA_BUCKETS.has(modelId)) return null;
+  if (
+    ANTIGRAVITY_DROPPED_QUOTA_BUCKETS.has(modelId) ||
+    ANTIGRAVITY_RETIRED_MODEL_IDS.has(modelId)
+  ) {
+    return null;
+  }
   return toClientAntigravityModelId(modelId);
 }