Updated inference endpoints added kimi model

app.py

@@ -14,6 +14,10 @@ load_dotenv()
 # #===========================================
 # updates = '''
 # Updates
+# + 02/06/2026
+# - Updated inference endpoints for HF models
+# - Added Kimi model 
+
 # + 01/10/2026
 # - Updated cooldown
 
@@ -169,30 +173,39 @@ if "last_request_time" not in st.session_state:
     
 
 model_links_hf ={
-      "Gemma-3-27B-it":{
+    "Gemma-3-27B-it":{
                       "inf_point":"https://router.huggingface.co/v1",
-                      "link":"google/gemma-3-27b-it-fast",
-                      },
-      "Meta-Llama-3.1-8B":{
-                      "inf_point":"https://router.huggingface.co/nebius/v1",
-                      "link":"meta-llama/Meta-Llama-3.1-8B-Instruct-fast",
-                      },
-      "Mistral-7B":{
-                      "inf_point":"https://router.huggingface.co/together/v1",
-                      "link":"mistralai/Mistral-7B-Instruct-v0.3",
+                      "link":"google/gemma-3-27b-it:scaleway",
                       },
-      "Gemma-2-27B-it":{
-                      "inf_point":"https://router.huggingface.co/nebius/v1",
-                      "link":"google/gemma-2-27b-it-fast",
-                      },
-      "Gemma-2-2B-it":{
-                      "inf_point":"https://router.huggingface.co/nebius/v1",
-                      "link":"google/gemma-2-2b-it-fast",
-                      },
-      "Zephyr-7B-β":{
-                      "inf_point":"https://router.huggingface.co/hf-inference/models/HuggingFaceH4/zephyr-7b-beta/v1",
-                      "link":"HuggingFaceH4/zephyr-7b-beta",
+    "Meta-Llama-3.1-8B":{
+                      "inf_point":"https://router.huggingface.co/v1",
+                      "link":"meta-llama/Meta-Llama-3.1-8B-Instruct:scaleway",
                       },
+    "DeepSeek-R1-Distill-Llama-70B":{
+                    "inf_point":"https://router.huggingface.co/v1",
+                    "link":"deepseek-ai/DeepSeek-R1-Distill-Llama-70B:scaleway",
+                     },
+    "Qwen2.5-Coder-32B-Instruct":{
+                        "inf_point":"https://router.huggingface.co/v1",
+                        "link":"Qwen/Qwen3-235B-A22B-Instruct-2507:scaleway",
+                        },
+
+      # "Mistral-7B":{
+      #                 "inf_point":"https://router.huggingface.co/v1",
+      #                 "link":"mistralai/Mistral-7B-Instruct-v0.2",
+      #                 },
+      # "Gemma-2-27B-it":{
+      #                 "inf_point":"https://router.huggingface.co/nebius/v1",
+      #                 "link":"google/gemma-2-27b-it-fast",
+      #                 },
+      # "Gemma-2-2B-it":{
+      #                 "inf_point":"https://router.huggingface.co/nebius/v1",
+      #                 "link":"google/gemma-2-2b-it-fast",
+      #                 },
+      # "Zephyr-7B-β":{
+      #                 "inf_point":"https://router.huggingface.co/hf-inference/models/HuggingFaceH4/zephyr-7b-beta/v1",
+      #                 "link":"HuggingFaceH4/zephyr-7b-beta",
+      #                 },
   }
 
 
@@ -205,6 +218,11 @@ model_links_groq ={
                       "inf_point":"https://api.groq.com/openai/v1",
                       "link":"llama-3.1-8b-instant",
                       },
+      "Kimi-K2-Instruct":{
+                  "inf_point":"https://api.groq.com/openai/v1",
+                  "link":"moonshotai/kimi-k2-instruct",
+                  },
+
       # "Gemma-2-9B-it":{
       #                 "inf_point":"https://api.groq.com/openai/v1",
       #                 "link":"gemma2-9b-it",
@@ -262,6 +280,22 @@ model_info ={
         {'description':"""The Llama (3.1) model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
             \nIt was created by the [**Meta's AI**](https://llama.meta.com/) team and has over  **8 billion parameters.** \n""",
         'logo':'Llama3_1_logo.png'},
+    "DeepSeek-R1-Distill-Llama-70B":
+        {'description':"""DeepSeek-R1-Distill-Llama-70B is a **Large Language Model (LLM)** distilled from the DeepSeek-R1 reasoning family using the Llama architecture. \
+            \nIt is designed to retain strong capabilities in reasoning, coding, and general text generation while being more accessible than the full DeepSeek-R1 model. \
+            \nLearn more on HuggingFace: https://huggingface.co/deepseek-ai/DeepSeek-R1-Distill-Llama-70B""",
+        'logo':'https://cdn-avatars.huggingface.co/v1/production/uploads/6538815d1bdb3c40db94fbfa/xMBly9PUMphrFVMxLX4kq.png'},
+    "Qwen2.5-Coder-32B-Instruct":
+        {'description':"""Qwen2.5-Coder-32B-Instruct is a **Large Language Model (LLM)** in the Qwen2.5-Coder series tailored for code generation, reasoning, and instruction-following tasks. \
+            \nBuilt on the Qwen2.5 architecture, this 32B-parameter model is optimized for coding, debugging, and developer use cases. \
+            \nLearn more on HuggingFace: https://huggingface.co/Qwen/Qwen2.5-Coder-32B-Instruct""",
+        'logo':'https://cdn-avatars.huggingface.co/v1/production/uploads/620760a26e3b7210c2ff1943/-s1gyJfvbE1RgO5iBeNOi.png'},
+    "Kimi-K2-Instruct":
+        {'description':"""The Kimi-K2-Instruct model is a **Large Language Model (LLM)** that's able to have question and answer interactions.\n \
+            \nIt was created by the [**Moonshot AI**](https://www.moonshot.cn/) team as part of the Kimi model family. \
+            \nThe model is designed for instruction following, reasoning, and general conversational tasks, with a strong focus on high-quality responses and long-context understanding.\n""",
+    'logo':'https://cdn-avatars.huggingface.co/v1/production/uploads/641c1e77c3983aa9490f8121/X1yT2rsaIbR9cdYGEVu0X.jpeg'},
+
 }