superbrain-server 1.0.43 → 1.0.45

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "superbrain-server",
-  "version": "1.0.43",
+  "version": "1.0.45",
   "description": "1-Line Auto-Installer and Server Execution wrapper for SuperBrain",
   "main": "index.js",
   "bin": {

package/payload/config/model_rankings.json

@@ -1,11 +1,11 @@
 {
   "groq_gpt_oss_20b": {
     "key": "groq_gpt_oss_20b",
-    "avg_response_s": 1.5378954618382221,
-    "success_count": 71,
+    "avg_response_s": 1.3635670783103586,
+    "success_count": 77,
     "fail_count": 6,
     "down_until": null,
-    "last_used": "2026-04-09T12:10:09.765726",
+    "last_used": "2026-04-09T17:35:12.529719",
     "last_error": null,
     "base_priority": 0.5
   },
@@ -173,49 +173,49 @@
     "key": "openrouter_llama33_70b",
     "avg_response_s": null,
     "success_count": 0,
-    "fail_count": 0,
-    "down_until": null,
+    "fail_count": 1,
+    "down_until": "2026-04-09T17:58:36.014066",
     "last_used": null,
-    "last_error": null,
+    "last_error": "429 rate limit: {\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"meta-llama/llama-3.3-70b-instruct:free is temporarily rate-limited upstream. Please retry shortly, or add yo",
     "base_priority": 7
   },
   "openrouter_deepseek_r1_0528": {
     "key": "openrouter_deepseek_r1_0528",
     "avg_response_s": null,
     "success_count": 0,
-    "fail_count": 0,
-    "down_until": null,
+    "fail_count": 2,
+    "down_until": "2026-04-09T17:42:02.945931",
     "last_used": null,
-    "last_error": null,
+    "last_error": "404 Client Error: Not Found for url: https://openrouter.ai/api/v1/chat/completions",
     "base_priority": 7.5
   },
   "openrouter_qwen3_235b": {
     "key": "openrouter_qwen3_235b",
     "avg_response_s": null,
     "success_count": 0,
-    "fail_count": 0,
-    "down_until": null,
+    "fail_count": 2,
+    "down_until": "2026-04-09T17:42:03.575625",
     "last_used": null,
-    "last_error": null,
+    "last_error": "404 Client Error: Not Found for url: https://openrouter.ai/api/v1/chat/completions",
     "base_priority": 8
   },
   "openrouter_hermes3_405b": {
     "key": "openrouter_hermes3_405b",
     "avg_response_s": null,
     "success_count": 0,
-    "fail_count": 0,
-    "down_until": null,
+    "fail_count": 1,
+    "down_until": "2026-04-09T17:58:37.929800",
     "last_used": null,
-    "last_error": null,
+    "last_error": "429 rate limit: {\"error\":{\"message\":\"Provider returned error\",\"code\":429,\"metadata\":{\"raw\":\"nousresearch/hermes-3-llama-3.1-405b:free is temporarily rate-limited upstream. Please retry shortly, or add",
     "base_priority": 8.5
   },
   "openrouter_gpt_oss_120b": {
     "key": "openrouter_gpt_oss_120b",
-    "avg_response_s": null,
-    "success_count": 0,
+    "avg_response_s": 3.9147901058197023,
+    "success_count": 2,
     "fail_count": 0,
     "down_until": null,
-    "last_used": null,
+    "last_used": "2026-04-09T17:37:11.262276",
     "last_error": null,
     "base_priority": 9
   },
@@ -323,8 +323,8 @@
     "key": "gemini_25_flash_lite_vision",
     "avg_response_s": 6.709401964075168,
     "success_count": 14,
-    "fail_count": 21,
-    "down_until": "2026-04-09T12:14:58.761896",
+    "fail_count": 24,
+    "down_until": "2026-04-09T17:40:03.986631",
     "last_used": "2026-02-24T09:45:09.831171",
     "last_error": "No module named 'google.generativeai'",
     "base_priority": 1.5
@@ -333,8 +333,8 @@
     "key": "gemini_25_pro_vision",
     "avg_response_s": null,
     "success_count": 0,
-    "fail_count": 24,
-    "down_until": "2026-04-09T12:14:58.758850",
+    "fail_count": 27,
+    "down_until": "2026-04-09T17:40:03.984360",
     "last_used": null,
     "last_error": "No module named 'google.generativeai'",
     "base_priority": 2
@@ -353,8 +353,8 @@
     "key": "gemini_3_pro_vision",
     "avg_response_s": null,
     "success_count": 0,
-    "fail_count": 23,
-    "down_until": "2026-04-09T12:14:58.760897",
+    "fail_count": 26,
+    "down_until": "2026-04-09T17:40:03.984360",
     "last_used": null,
     "last_error": "No module named 'google.generativeai'",
     "base_priority": 3
@@ -363,8 +363,8 @@
     "key": "gemini_31_pro_vision",
     "avg_response_s": null,
     "success_count": 0,
-    "fail_count": 22,
-    "down_until": "2026-04-09T12:14:58.762895",
+    "fail_count": 25,
+    "down_until": "2026-04-09T17:40:03.987716",
     "last_used": null,
     "last_error": "No module named 'google.generativeai'",
     "base_priority": 3.5
@@ -373,8 +373,8 @@
     "key": "gemini_20_flash_vision",
     "avg_response_s": null,
     "success_count": 0,
-    "fail_count": 21,
-    "down_until": "2026-04-09T12:14:58.764895",
+    "fail_count": 24,
+    "down_until": "2026-04-09T17:40:03.987716",
     "last_used": null,
     "last_error": "No module named 'google.generativeai'",
     "base_priority": 4
@@ -383,8 +383,8 @@
     "key": "gemini_20_flash_lite_vision",
     "avg_response_s": null,
     "success_count": 0,
-    "fail_count": 21,
-    "down_until": "2026-04-09T12:14:58.765718",
+    "fail_count": 24,
+    "down_until": "2026-04-09T17:40:03.987716",
     "last_used": null,
     "last_error": "No module named 'google.generativeai'",
     "base_priority": 4.5
@@ -393,19 +393,19 @@
     "key": "gemini_15_flash_vision",
     "avg_response_s": null,
     "success_count": 0,
-    "fail_count": 21,
-    "down_until": "2026-04-09T12:14:58.765718",
+    "fail_count": 24,
+    "down_until": "2026-04-09T17:40:03.987716",
     "last_used": null,
     "last_error": "No module named 'google.generativeai'",
     "base_priority": 4.8
   },
   "groq_llama4_scout_vision": {
     "key": "groq_llama4_scout_vision",
-    "avg_response_s": 1.7195895721505887,
-    "success_count": 46,
+    "avg_response_s": 1.5534682586944504,
+    "success_count": 52,
     "fail_count": 5,
     "down_until": null,
-    "last_used": "2026-04-09T12:10:03.943317",
+    "last_used": "2026-04-09T17:35:07.357177",
     "last_error": null,
     "base_priority": 5
   },

package/payload/core/model_router.py

@@ -489,6 +489,9 @@ def _has_image_input(m: Dict) -> bool:
     return "image" in str(mods)
 
 
+class RateLimitError(Exception):
+    pass
+
 # ─────────────────────────────────────────────────────────────────────────────
 #  MODEL ROUTER
 # ─────────────────────────────────────────────────────────────────────────────
@@ -707,10 +710,10 @@ class ModelRouter:
             resp.raise_for_status()
             all_models = resp.json().get("data", [])
         except Exception as e:
-            if "429" in str(e) or "quota" in str(e).lower():
-                raise RateLimitError("Quota limit hit")
-            raise e
-            print(f"⚠️  OpenRouter model discovery failed: {e}")
+            # if "429" in str(e) or "quota" in str(e).lower():
+            #     raise RateLimitError("Quota limit hit")
+            # raise e
+            print(f"⚠️  OpenRouter free model discovery failed: {e}")
             return
 
         # Filter for free models (pricing.prompt == 0 or :free suffix)
@@ -1103,8 +1106,9 @@ class ModelRouter:
                 return result
 
             except Exception as e:
-                if "429" in str(e) or "quota" in str(e).lower():
-                    raise RateLimitError("Quota limit hit")
+                # Do not immediately abort on quota, try next model
+                # if "429" in str(e) or "quota" in str(e).lower():
+                #     raise RateLimitError("Quota limit hit")
                 status = 429 if "429" in str(e) else 0
                 self._record_failure(key, str(e), status_code=status)
                 print(f"  ✗ Failed ({type(e).__name__}), trying next …", flush=True)
@@ -1149,8 +1153,9 @@ class ModelRouter:
                 return result
 
             except Exception as e:
-                if "429" in str(e) or "quota" in str(e).lower():
-                    raise RateLimitError("Quota limit hit")
+                # Do not immediately abort on quota, try next model
+                # if "429" in str(e) or "quota" in str(e).lower():
+                #     raise RateLimitError("Quota limit hit")
                 status = 429 if "429" in str(e) else 0
                 self._record_failure(key, str(e), status_code=status)
                 print(f"  ✗ Failed ({type(e).__name__}), trying next …", flush=True)

package/payload/start.py

@@ -410,6 +410,12 @@ OLLAMA_MODEL = "qwen3-vl:4b"   # vision-language model, fits ~6 GB VRAM / ~8 GB
 def setup_ollama():
     h1("Step 4 of 7 — Offline AI Model (Ollama)")
 
+    machine = platform.machine().lower()
+    is_arm = any(arch in machine for arch in ["arm", "aarch64"])
+    if is_arm:
+        info(f"ARM device detected ({machine}) — skipping Ollama local installation (heavy resource/unsupported).")
+        return
+
     keys = _load_saved_api_keys()
     has_cloud_key = any(keys.get(k) for k in ("GEMINI_API_KEY", "GROQ_API_KEY", "OPENROUTER_API_KEY"))
 
@@ -547,6 +553,12 @@ WHISPER_MODELS = {
 def setup_whisper():
     h1("Step 5 of 7 — Offline Audio Transcription (Whisper)")
 
+    machine = platform.machine().lower()
+    is_arm = any(arch in machine for arch in ["arm", "aarch64"])
+    if is_arm:
+        info(f"ARM device detected ({machine}) — skipping local Whisper installation to prevent wheel build failures.")
+        return
+
     keys = _load_saved_api_keys()
     has_groq_key = bool(keys.get("GROQ_API_KEY") or os.getenv("GROQ_API_KEY"))