feat: add tiyuvta as an OpenAI-compatible provider

Registry entry plus the provider_endpoints_support.json row that code-quality
requires for every openai_like provider.

No param_mappings: the endpoint already accepts max_completion_tokens as an alias
for max_tokens, so the mapping other entries carry would be a no-op here.

Only chat_completions is claimed in the support matrix. The endpoint serves chat
completions and legacy completions; the matrix has no field for the latter, and
claiming messages or responses would assert LiteLLM's bridge works here without
having run it.

litellm/llms/openai_like/providers.json

@@ -175,6 +175,12 @@
     "base_class": "openai_gpt",
     "supported_endpoints": ["/v1/chat/completions", "/v1/responses", "/v1/messages"]
   },
+  "tiyuvta": {
+    "base_url": "https://api.tiyuvta.ai/v1",
+    "api_key_env": "TIYUVTA_API_KEY",
+    "api_base_env": "TIYUVTA_API_BASE",
+    "supported_endpoints": ["/v1/chat/completions", "/v1/completions"]
+  },
   "pinstripes": {
     "base_url": "https://pinstripes.io/v1",
     "api_key_env": "PINSTRIPES_API_KEY",

provider_endpoints_support.json

@@ -2366,6 +2366,23 @@
         "interactions": true
       }
     },
+    "tiyuvta": {
+      "display_name": "tiyuvta (`tiyuvta`)",
+      "url": "https://inference.tiyuvta.ai/docs",
+      "endpoints": {
+        "chat_completions": true,
+        "messages": false,
+        "responses": false,
+        "embeddings": false,
+        "image_generations": false,
+        "audio_transcriptions": false,
+        "audio_speech": false,
+        "moderations": false,
+        "batches": false,
+        "rerank": false,
+        "a2a": false
+      }
+    },
     "together_ai": {
       "display_name": "Together AI (`together_ai`)",
       "url": "https://docs.litellm.ai/docs/providers/togetherai",