diff --git a/src/modules/llms/server/openai/models/groq.models.ts b/src/modules/llms/server/openai/models/groq.models.ts
index 161bdc5f1..7c3522872 100644
--- a/src/modules/llms/server/openai/models/groq.models.ts
+++ b/src/modules/llms/server/openai/models/groq.models.ts
@@ -14,7 +14,7 @@ const DEV_DEBUG_GROQ_MODELS = Release.IsNodeDevBuild; // not in staging to reduc
  * Groq models.
  * - models list: https://console.groq.com/docs/models
  * - pricing: https://groq.com/pricing/
- * - updated: 2026-01-30
+ * - updated: 2026-02-18
  */
 const _knownGroqModels: ManualMappings = [
 
@@ -59,20 +59,10 @@ const _knownGroqModels: ManualMappings = [
     interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn],
     chatPrice: { input: 1.00, output: 3.00 },
   },
-  {
-    isLegacy: true,
-    idPrefix: 'moonshotai/kimi-k2-instruct',
-    label: 'Kimi K2 Instruct (Deprecated)',
-    description: 'Deprecated on 2025-10-10, redirects to kimi-k2-instruct-0905.',
-    contextWindow: 131072,
-    maxCompletionTokens: 16384,
-    interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn],
-    chatPrice: { input: 1.00, output: 3.00 },
-    hidden: true,
-  },
   // REMOVED MODELS (no longer returned by API):
   // - (Jan 21, 2026) qwen-qwq-32b, qwen-2.5-32b, qwen-2.5-coder-32b
   // - (Jan 21, 2026) deepseek-r1-distill-llama-70b, deepseek-r1-distill-qwen-32b
+  // - (Feb 18, 2026) moonshotai/kimi-k2-instruct (deprecated redirect, removed from docs)
 
 
   // Production Models - Compound Systems (pass-through pricing to underlying models)
@@ -106,8 +96,9 @@ const _knownGroqModels: ManualMappings = [
     chatPrice: { input: 0.15, output: 0.60 },
   },
   {
+    isPreview: true,
     idPrefix: 'openai/gpt-oss-safeguard-20b',
-    label: 'GPT OSS Safeguard 20B',
+    label: 'GPT OSS Safeguard 20B (Preview)',
     description: 'OpenAI safety classification model (20B MoE). Purpose-built for content moderation with Harmony response format. 131K context, 65K max output. ~1000 t/s on Groq.',
     contextWindow: 131072,
     maxCompletionTokens: 65536,
@@ -125,15 +116,7 @@ const _knownGroqModels: ManualMappings = [
   },
 
   // Production Models - Meta
-  {
-    idPrefix: 'meta-llama/llama-guard-4-12b',
-    label: 'Llama Guard 4 · 12B',
-    description: 'Meta multimodal content moderation (12B params). Classifies text and images. 131K context, 1K max output. ~1200 t/s on Groq.',
-    contextWindow: 131072,
-    maxCompletionTokens: 1024,
-    interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn],
-    chatPrice: { input: 0.20, output: 0.20 },
-  },
+  // (Feb 18, 2026) meta-llama/llama-guard-4-12b removed from docs
   {
     idPrefix: 'llama-3.3-70b-versatile',
     label: 'Llama 3.3 · 70B Versatile',
@@ -153,17 +136,7 @@ const _knownGroqModels: ManualMappings = [
     chatPrice: { input: 0.05, output: 0.08 },
   },
 
-  // Preview Models - SDAIA
-  {
-    isPreview: true,
-    idPrefix: 'allam-2-7b',
-    label: 'ALLaM 2 7B (Preview)',
-    description: 'ALLaM 2 7B by Saudi Data and AI Authority (SDAIA). Bilingual Arabic-English model. 4K context and max output.',
-    contextWindow: 4096,
-    maxCompletionTokens: 4096,
-    interfaces: [LLM_IF_OAI_Chat, LLM_IF_OAI_Fn],
-    hidden: true, // Pricing unknown
-  },
+  // (Feb 18, 2026) allam-2-7b (SDAIA) removed from docs and pricing
 
 ];