chore(llm): update callers for renamed model variants

Picks up the registry rename in the front-end and Tauri command layer:

- src/lib/types/app.ts: LlmModelIdStr now lists the four new ids
  (qwen3_5_2b / qwen3_5_4b / qwen3_5_9b / qwen3_6_27b).
- src/lib/pages/SettingsPage.svelte: LLM_MODELS table rebuilt with
  four tiers (Minimal / Standard / High / Maximum), matching subtitles
  and download-size copy. selectedLlmModelId fallback, hardware-warning
  thresholds, tier-availability check, and ensureRecommendedLlmTier
  fallback all retargeted at the new ids. The Maximum tier surfaces a
  64 GB / 24 GB warning so users with mid-range hardware see honest
  expectations.
- src-tauri/src/commands/llm.rs and commands/tasks.rs: doc-comment
  examples refreshed (Qwen3 4B → Qwen3.5 4B, Qwen3's tokenizer →
  Qwen's tokenizer — the BPE family is shared).
- src/lib/stores/llmStatus.svelte.ts: chip-detail example updated.

cargo build --workspace clean. cargo test --workspace clean.
npx svelte-check reports one pre-existing error in vite.config.js
(unused @ts-expect-error directive, dates back to the original
scaffold commit 9926a42); not introduced here, out of scope to fix.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

src-tauri/src/commands/llm.rs

@@ -159,7 +159,7 @@ pub struct LlmTestResult {
     pub category: String,
     /// `true` when the LLM is healthy and usable after the test.
     pub ok: bool,
-    /// One-line status copy for the Settings chip ("Qwen3 4B ready").
+    /// One-line status copy for the Settings chip ("Qwen3.5 4B ready").
     pub message: String,
     /// Optional actionable next step ("Click Download", "Delete and
     /// re-download", "Pick a smaller tier"). Absent when the state is

src-tauri/src/commands/tasks.rs

@@ -250,7 +250,7 @@ fn to_llm_examples(rows: Vec<FeedbackRow>) -> Vec<LlmFeedbackExample> {
         .collect()
 }
 
-/// Rough character budget for the few-shot block. Qwen3's tokenizer
+/// Rough character budget for the few-shot block. Qwen's tokenizer
 /// averages ~3.5 chars per token in English, so 2000 chars is ~570
 /// tokens — well inside the 64-token reserve + response-token gap
 /// against the 8192-token context cap (see `LlmEngine::generate`).

src/lib/pages/SettingsPage.svelte

@@ -53,25 +53,32 @@
 
   const LLM_MODELS = [
     {
-      id: "qwen3_1_7b",
-      label: "Low",
-      subtitle: "Qwen3 1.7B",
+      id: "qwen3_5_2b",
+      label: "Minimal",
+      subtitle: "Qwen3.5 2B",
       fit: "8 GB RAM, CPU-heavy machines",
-      size: "~1.1 GB",
+      size: "~1.3 GB",
     },
     {
-      id: "qwen3_4b_instruct_2507",
-      label: "Default",
-      subtitle: "Qwen3 4B Instruct 2507",
-      fit: "16 GB RAM or 8 GB+ VRAM",
-      size: "~2.5 GB",
+      id: "qwen3_5_4b",
+      label: "Standard",
+      subtitle: "Qwen3.5 4B",
+      fit: "16 GB RAM or 6 GB+ VRAM",
+      size: "~2.7 GB",
     },
     {
-      id: "qwen3_14b",
+      id: "qwen3_5_9b",
       label: "High",
-      subtitle: "Qwen3 14B",
-      fit: "32 GB RAM or 16 GB+ VRAM",
-      size: "~10.5 GB",
+      subtitle: "Qwen3.5 9B",
+      fit: "32 GB RAM and 12 GB+ VRAM",
+      size: "~5.7 GB",
+    },
+    {
+      id: "qwen3_6_27b",
+      label: "Maximum",
+      subtitle: "Qwen3.6 27B",
+      fit: "64 GB RAM and 24 GB+ VRAM",
+      size: "~17 GB",
     },
   ];
 
@@ -459,7 +466,7 @@
   }
 
   function selectedLlmModelId() {
-    return settings.llmModelId || "qwen3_4b_instruct_2507";
+    return settings.llmModelId || "qwen3_5_4b";
   }
 
   function llmModelStatus(modelId) {
@@ -476,19 +483,23 @@
 
   function llmHardwareWarning(modelId) {
     const ramMb = systemInfo?.ram_mb || 0;
-    if (modelId === "qwen3_14b" && ramMb < 32768) {
+    if (modelId === "qwen3_6_27b" && ramMb < 65536) {
+      return "Maximum tier needs 64 GB RAM (or a 24 GB GPU) to avoid heavy swap. Expect slow responses on this machine.";
+    }
+    if (modelId === "qwen3_5_9b" && ramMb < 32768) {
       return "High tier will swap heavily on this machine. Expect slow responses.";
     }
-    if (modelId === "qwen3_4b_instruct_2507" && ramMb < 16384 && !hasGpuAcceleration()) {
-      return "Default tier is best with 16 GB RAM or a GPU-backed build.";
+    if (modelId === "qwen3_5_4b" && ramMb < 16384 && !hasGpuAcceleration()) {
+      return "Standard tier is best with 16 GB RAM or a GPU-backed build.";
     }
     return "";
   }
 
   function llmTierAvailable(modelId) {
     const ramMb = systemInfo?.ram_mb || 0;
-    if (modelId === "qwen3_14b") return ramMb >= 32768;
-    if (modelId === "qwen3_4b_instruct_2507") return ramMb >= 16384 || hasGpuAcceleration();
+    if (modelId === "qwen3_6_27b") return ramMb >= 65536;
+    if (modelId === "qwen3_5_9b") return ramMb >= 32768;
+    if (modelId === "qwen3_5_4b") return ramMb >= 16384 || hasGpuAcceleration();
     return true;
   }
 
@@ -497,7 +508,7 @@
     try {
       settings.llmModelId = await invoke("recommend_llm_tier");
     } catch {
-      settings.llmModelId = "qwen3_4b_instruct_2507";
+      settings.llmModelId = "qwen3_5_4b";
     }
   }
 
@@ -1732,7 +1743,7 @@
               <p class="text-[11px] text-text-tertiary mt-3">
                 Recommended for this machine:
                 <span class="text-text">
-                  {LLM_MODELS.find((model) => model.id === (settings.llmModelId || "qwen3_4b_instruct_2507"))?.subtitle || "Qwen3 4B Instruct 2507"}
+                  {LLM_MODELS.find((model) => model.id === (settings.llmModelId || "qwen3_5_4b"))?.subtitle || "Qwen3.5 4B"}
                 </span>
                 {#if systemInfo}
                   · {Math.round((systemInfo.ram_mb || 0) / 1024)} GB RAM detected

src/lib/stores/llmStatus.svelte.ts

@@ -11,7 +11,7 @@ export type LlmStatusKind = "off" | "warming" | "ready" | "generating" | "error"
 
 export interface LlmStatusState {
   kind: LlmStatusKind;
-  /// Optional short phrase for the chip to surface (e.g. "Loading Qwen3 4B").
+  /// Optional short phrase for the chip to surface (e.g. "Loading Qwen3.5 4B").
   detail: string | null;
 }
 

src/lib/types/app.ts

@@ -11,7 +11,11 @@ export type WhisperModelSize =
   | "Medium"
   | "Distil-L";
 export type AiTier = "off" | "cleanup" | "tasks";
-export type LlmModelIdStr = "qwen3_1_7b" | "qwen3_4b_instruct_2507" | "qwen3_14b";
+export type LlmModelIdStr =
+  | "qwen3_5_2b"
+  | "qwen3_5_4b"
+  | "qwen3_5_9b"
+  | "qwen3_6_27b";
 export type LlmPromptPreset = "default" | "email" | "notes" | "code";
 export type AiGpuConcurrency = "parallel" | "sequential";
 export type TaskBucket = "inbox" | "today" | "soon" | "later";