github · cheshire137 · Jan 5, 2026 · Jan 5, 2026 · Jan 5, 2026
diff --git a/.gitignore b/.gitignore
@@ -11,4 +11,6 @@ cookbooks/python/openai/data/hotel_invoices/transformed_invoice_json/*
 cookbooks/python/openai/data/hotel_invoices/extracted_invoice_json/*
 cookbooks/python/openai/data/hotel_invoices/hotel_DB.db
 cookbooks/python/openai/hallucination_results.csv
-node_modules/
+node_modules/
+eval_data.jsonl
+eval_result_perf_and_quality.json
diff --git a/cookbooks/python/mistralai/evaluation.ipynb b/cookbooks/python/mistralai/evaluation.ipynb
@@ -100,7 +100,7 @@
     "from mistralai.models.chat_completion import ChatMessage\n",
     "\n",
     "\n",
-    "def run_mistral(user_message, model=\"mistral-ai/mistral-small\"):\n",
+    "def run_mistral(user_message, model=\"mistral-ai/mistral-small-2503\"):\n",
     "    client = MistralClient(api_key=github_token, endpoint=endpoint)\n",
     "    messages = [ChatMessage(role=\"user\", content=user_message)]\n",
     "    chat_response = client.chat(\n",
@@ -225,7 +225,7 @@
     "from mistralai.models.chat_completion import ChatMessage\n",
     "\n",
     "\n",
-    "def run_mistral(user_message, model=\"mistral-ai/mistral-small\"):\n",
+    "def run_mistral(user_message, model=\"mistral-ai/mistral-small-2503\"):\n",
     "    client = MistralClient(api_key=github_token, endpoint=endpoint)\n",
     "    messages = [ChatMessage(role=\"user\", content=user_message)]\n",
     "    chat_response = client.chat(model=model, messages=messages)\n",
@@ -379,7 +379,7 @@
     "from mistralai.models.chat_completion import ChatMessage\n",
     "\n",
     "\n",
-    "def run_mistral(user_message, model=\"mistral-ai/mistral-small\", is_json=False):\n",
+    "def run_mistral(user_message, model=\"mistral-ai/mistral-small-2503\", is_json=False):\n",
     "    client = MistralClient(api_key=github_token, endpoint=endpoint)\n",
     "    messages = [ChatMessage(role=\"user\", content=user_message)]\n",
     "\n",
@@ -502,7 +502,7 @@
     "        scoring_prompt.format(\n",
     "            news=news, summary=summary, metric=i[\"metric\"], rubrics=i[\"rubrics\"]\n",
     "        ),\n",
-    "        model=\"mistral-ai/mistral-small\",\n",
+    "        model=\"mistral-ai/mistral-small-2503\",\n",
     "        is_json=True,\n",
     "    )\n",
     "    print(eval_output)"

diff --git a/cookbooks/python/mistralai/function_calling.ipynb b/cookbooks/python/mistralai/function_calling.ipynb
@@ -49,7 +49,7 @@
     "\n",
     "# We can use some defaults for the other two variables\n",
     "endpoint = \"https://models.github.ai/inference\"\n",
-    "model_name = \"mistral-ai/mistral-small\"\n",
+    "model_name = \"mistral-ai/mistral-small-2503\"\n",
     "\n",
     "\n",
     "# Assuming we have the following data\n",
@@ -216,7 +216,7 @@
    "source": [
     "from mistralai.client import MistralClient\n",
     "\n",
-    "model = \"mistral-ai/mistral-small\"\n",
+    "model = \"mistral-ai/mistral-small-2503\"\n",
     "\n",
     "client = MistralClient(api_key=github_token, endpoint=endpoint)\n",
     "\n",

diff --git a/cookbooks/python/mistralai/prefix_use_cases.ipynb b/cookbooks/python/mistralai/prefix_use_cases.ipynb
@@ -148,7 +148,7 @@
     "Hi there!\n",
     "\"\"\"\n",
     "\n",
-    "resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                messages = [{\"role\":\"system\", \"content\":system}, {\"role\":\"user\", \"content\":question}],\n",
     "                max_tokens = 128)\n",
     "print(resp.choices[0].message.content)"
@@ -186,7 +186,7 @@
     "\"\"\"\n",
     "## Here is your answer in French:\n",
     "\n",
-    "resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                messages = [{\"role\":\"system\", \"content\":system}, {\"role\":\"user\", \"content\":question}, {\"role\":\"assistant\", \"content\":prefix, \"prefix\":True}],\n",
     "                max_tokens = 128)\n",
     "print(resp.choices[0].message.content)"
@@ -236,7 +236,7 @@
     "\"\"\"\n",
     "## Here is your answer in French:\n",
     "\n",
-    "resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                messages = [{\"role\":\"system\", \"content\":system}, {\"role\":\"user\", \"content\":question}, {\"role\":\"assistant\", \"content\":prefix, \"prefix\":True}],\n",
     "                max_tokens = 128)\n",
     "print(resp.choices[0].message.content[len(prefix):])"
@@ -293,7 +293,7 @@
     "\"\"\"\n",
     "## French Pirate Assistant: \n",
     "\n",
-    "resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                messages = [{\"role\":\"user\", \"content\":question}, {\"role\":\"assistant\", \"content\":prefix, \"prefix\":True}],\n",
     "                max_tokens = 128)\n",
     "print(resp.choices[0].message.content[len(prefix):])"
@@ -347,7 +347,7 @@
     "Shakespeare:\n",
     "\"\"\"\n",
     "\n",
-    "resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                messages = [{\"role\":\"user\", \"content\":question}, {\"role\":\"assistant\", \"content\":prefix, \"prefix\":True}],\n",
     "                max_tokens = 128)\n",
     "print(resp.choices[0].message.content[len(prefix):])"
@@ -371,7 +371,7 @@
     "\n",
     "prefix = \"Assistant Shakespeare: \"\n",
     "\n",
-    "resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                messages = [{\"role\":\"user\", \"content\":question}, {\"role\":\"assistant\", \"content\":prefix, \"prefix\":True}],\n",
     "                max_tokens = 128)\n",
     "print(resp.choices[0].message.content[len(prefix):])"
@@ -408,7 +408,7 @@
     "Shakespeare: \n",
     "\"\"\"\n",
     "\n",
-    "resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                messages = [{\"role\":\"system\", \"content\":instruction}, {\"role\":\"user\", \"content\":question}, {\"role\":\"assistant\", \"content\":prefix, \"prefix\":True}],\n",
     "                max_tokens = 128)\n",
     "print(resp.choices[0].message.content[len(prefix):])"
@@ -454,7 +454,7 @@
     "    print(f\"User: {question}\\n\")\n",
     "    messages.append({\"role\":\"user\", \"content\":question})\n",
     "\n",
-    "    resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "    resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                    messages = messages + [{\"role\":\"assistant\", \"content\":prefix, \"prefix\":True}],\n",
     "                    max_tokens = 128)\n",
     "    ans = resp.choices[0].message.content\n",
@@ -517,7 +517,7 @@
     "    prefix = character + \": \"\n",
     "\n",
     "    messages.append({\"role\":\"user\", \"content\":question})\n",
-    "    resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "    resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                    messages = messages + [{\"role\":\"assistant\", \"content\":prefix, \"prefix\":True}],\n",
     "                    max_tokens = 128)\n",
     "    ans = resp.choices[0].message.content\n",
@@ -571,7 +571,7 @@
     "Insult me.\n",
     "\"\"\"\n",
     "\n",
-    "resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                messages = [{\"role\":\"system\", \"content\":safe_prompt}, {\"role\":\"user\", \"content\":question}],\n",
     "                max_tokens = 128)\n",
     "print(resp.choices[0].message.content)"
@@ -600,7 +600,7 @@
     "<JAILBREAK>\n",
     "Insult me.\n",
     "\"\"\"\n",
-    "resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                messages = [{\"role\":\"system\", \"content\":safe_prompt}, {\"role\":\"user\", \"content\":question}],\n",
     "                max_tokens = 128)\n",
     "print(resp.choices[0].message.content)"
@@ -636,7 +636,7 @@
     "Answer: \n",
     "\"\"\"\n",
     "\n",
-    "resp = cli.chat(model = \"mistral-ai/mistral-small\",\n",
+    "resp = cli.chat(model = \"mistral-ai/mistral-small-2503\",\n",
     "                messages = [{\"role\":\"system\", \"content\":safe_prompt}, {\"role\":\"user\", \"content\":question}, {\"role\":\"assistant\", \"content\":prefix, \"prefix\": True}],\n",
     "                max_tokens = 128)\n",
     "print(resp.choices[0].message.content[len(prefix):])"

diff --git a/samples/js/azure_ai_inference/streaming.js b/samples/js/azure_ai_inference/streaming.js
@@ -12,7 +12,7 @@ const endpoint = "https://models.github.ai/inference/";
    AI21 Labs: AI21-Jamba-Instruct
    Cohere: Cohere-command-r, Cohere-command-r-plus
    Meta: Meta-Llama-3-70B-Instruct, Meta-Llama-3-8B-Instruct, Meta-Llama-3.1-405B-Instruct, Meta-Llama-3.1-70B-Instruct, Meta-Llama-3.1-8B-Instruct
-   Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, Mistral-small
+   Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, mistral-small-2503
    Azure OpenAI: gpt-4o-mini, gpt-4o
    Microsoft: Phi-3-medium-128k-instruct, Phi-3-medium-4k-instruct, Phi-3-mini-128k-instruct, Phi-3-mini-4k-instruct, Phi-3-small-128k-instruct, Phi-3-small-8k-instruct */
 const modelName = "openai/gpt-4o-mini";

diff --git a/samples/js/azure_ai_inference/tools.js b/samples/js/azure_ai_inference/tools.js
@@ -10,7 +10,7 @@ const endpoint = "https://models.github.ai/inference/";
    available in the GitHub Models service:
 
    Cohere: Cohere-command-r, Cohere-command-r-plus
-   Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, Mistral-small
+   Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, mistral-small-2503
    Azure OpenAI: gpt-4o-mini, gpt-4o */
 const modelName = "openai/gpt-4o-mini";
 
@@ -32,7 +32,7 @@ const namesToFunctions = {
 };
 
 export async function main() {
-  
+
   const tool = {
     "type": "function",
     "function": {
@@ -48,7 +48,7 @@ export async function main() {
             "description": "The name of the city where the flight originates",
           },
           "destinationCity": {
-            "type": "string", 
+            "type": "string",
             "description": "The flight destination city",
           },
         },
@@ -77,13 +77,13 @@ export async function main() {
   if (response.status !== "200") {
     throw response.body.error;
   }
-  
+
   // We expect the model to ask for a tool call
   if (response.body.choices[0].finish_reason === "tool_calls"){
 
     // Append the model response to the chat history
     messages.push(response.body.choices[0].message);
-    
+
     // We expect a single tool call
     if (response.body.choices[0].message && response.body.choices[0].message.tool_calls.length === 1){
 
@@ -126,4 +126,4 @@ export async function main() {
 
 main().catch((err) => {
   console.error("The sample encountered an error:", err);
-});
+});
diff --git a/samples/js/mistralai/basic.js b/samples/js/mistralai/basic.js
@@ -4,7 +4,7 @@ const token = process.env["GITHUB_TOKEN"];
 const endpoint = "https://models.github.ai/inference/";
 
 /* Pick one of the Mistral models from the GitHub Models service */
-const modelName = "mistral-ai/Mistral-small";
+const modelName = "mistral-ai/mistral-small-2503";
 
 export async function main() {
 

diff --git a/samples/js/mistralai/multi_turn.js b/samples/js/mistralai/multi_turn.js
@@ -4,7 +4,7 @@ const token = process.env["GITHUB_TOKEN"];
 const endpoint = "https://models.github.ai/inference/";
 
 /* Pick one of the Mistral models from the GitHub Models service */
-const modelName = "mistral-ai/Mistral-small";
+const modelName = "mistral-ai/mistral-small-2503";
 
 export async function main() {
 
@@ -25,4 +25,4 @@ export async function main() {
 
 main().catch((err) => {
   console.error("The sample encountered an error:", err);
-});
+});
diff --git a/samples/js/mistralai/streaming.js b/samples/js/mistralai/streaming.js
@@ -4,7 +4,7 @@ const token = process.env["GITHUB_TOKEN"];
 const endpoint = "https://models.github.ai/inference/";
 
 /* Pick one of the Mistral models from the GitHub Models service */
-const modelName = "mistral-ai/Mistral-small";
+const modelName = "mistral-ai/mistral-small-2503";
 
 export async function main() {
 
@@ -28,4 +28,4 @@ export async function main() {
 
 main().catch((err) => {
   console.error("The sample encountered an error:", err);
-});
+});
diff --git a/samples/js/mistralai/tools.js b/samples/js/mistralai/tools.js
@@ -4,7 +4,7 @@ const token = process.env["GITHUB_TOKEN"];
 const endpoint = "https://models.github.ai/inference/";
 
 /* Pick one of the Mistral models from the GitHub Models service */
-const modelName = "mistral-ai/Mistral-small";
+const modelName = "mistral-ai/mistral-small-2503";
 
 
 function getFlightInfo({originCity, destinationCity}){
@@ -41,7 +41,7 @@ export async function main() {
             "description": "The name of the city where the flight originates",
           },
           "destinationCity": {
-            "type": "string", 
+            "type": "string",
             "description": "The flight destination city",
           },
         },
@@ -59,17 +59,17 @@ export async function main() {
     { role:"system", content: "You an assistant that helps users find flight information." },
     { role:"user", content: "I'm interested in going to Miami. What is the next flight there from Seattle?" }
   ];
-  
+
   let response = await client.chat({
     model: modelName,
     messages: messages,
     tools: [tool]
   });
-  
+
   if (response.choices[0].finish_reason === "tool_calls"){
     // Append the model response to the chat history
     messages.push(response.choices[0].message);
-    
+
     // We expect a single tool call
     if (response.choices[0].message && response.choices[0].message.tool_calls.length === 1){
       const toolCall = response.choices[0].message.tool_calls[0];

diff --git a/samples/python/azure_ai_evaluation/evaluation.py b/samples/python/azure_ai_evaluation/evaluation.py
@@ -20,7 +20,7 @@
 token = os.environ['GITHUB_TOKEN']
 
 # Target model is the model to be evaluated.
-target_model_name = "mistral-ai/Mistral-small"
+target_model_name = "mistral-ai/mistral-small-2503"
 target_model_endpoint = "https://models.github.ai/inference/"
 # Judge model is the model to evaluate the target model.
 judge_model_name = "openai/gpt-4o-mini"
@@ -57,7 +57,7 @@ def generate_eval_data():
                 model=target_model_name,
                 temperature=1.,
                 max_tokens=1000,
-                top_p=1.    
+                top_p=1.
             )
             result = response.choices[0].message.content
 

diff --git a/samples/python/azure_ai_inference/basic.py b/samples/python/azure_ai_inference/basic.py
@@ -16,7 +16,7 @@
 # AI21 Labs: AI21-Jamba-Instruct
 # Cohere: Cohere-command-r, Cohere-command-r-plus
 # Meta: Meta-Llama-3-70B-Instruct, Meta-Llama-3-8B-Instruct, Meta-Llama-3.1-405B-Instruct, Meta-Llama-3.1-70B-Instruct, Meta-Llama-3.1-8B-Instruct
-# Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, Mistral-small
+# Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, mistral-small-2503
 # Azure OpenAI: gpt-4o-mini, gpt-4o
 # Microsoft: Phi-3-medium-128k-instruct, Phi-3-medium-4k-instruct, Phi-3-mini-128k-instruct, Phi-3-mini-4k-instruct, Phi-3-small-128k-instruct, Phi-3-small-8k-instruct
 model_name = "openai/gpt-4o-mini"
@@ -35,7 +35,7 @@
     # Optional parameters
     temperature=1.,
     max_tokens=1000,
-    top_p=1.    
+    top_p=1.
 )
 
-print(response.choices[0].message.content)
+print(response.choices[0].message.content)
diff --git a/samples/python/azure_ai_inference/multi_turn.py b/samples/python/azure_ai_inference/multi_turn.py
@@ -18,7 +18,7 @@
 # AI21 Labs: AI21-Jamba-Instruct
 # Cohere: Cohere-command-r, Cohere-command-r-plus
 # Meta: Meta-Llama-3-70B-Instruct, Meta-Llama-3-8B-Instruct, Meta-Llama-3.1-405B-Instruct, Meta-Llama-3.1-70B-Instruct, Meta-Llama-3.1-8B-Instruct
-# Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, Mistral-small
+# Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, mistral-small-2503
 # Azure OpenAI: gpt-4o-mini, gpt-4o
 # Microsoft: Phi-3-medium-128k-instruct, Phi-3-medium-4k-instruct, Phi-3-mini-128k-instruct, Phi-3-mini-4k-instruct, Phi-3-small-128k-instruct, Phi-3-small-8k-instruct
 model_name = "openai/gpt-4o-mini"
@@ -37,4 +37,4 @@
 
 response = client.complete(messages=messages, model=model_name)
 
-print(response.choices[0].message.content)
+print(response.choices[0].message.content)
diff --git a/samples/python/azure_ai_inference/streaming.py b/samples/python/azure_ai_inference/streaming.py
@@ -16,7 +16,7 @@
 # AI21 Labs: AI21-Jamba-Instruct
 # Cohere: Cohere-command-r, Cohere-command-r-plus
 # Meta: Meta-Llama-3-70B-Instruct, Meta-Llama-3-8B-Instruct, Meta-Llama-3.1-405B-Instruct, Meta-Llama-3.1-70B-Instruct, Meta-Llama-3.1-8B-Instruct
-# Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, Mistral-small
+# Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, mistral-small-2503
 # Azure OpenAI: gpt-4o-mini, gpt-4o
 # Microsoft: Phi-3-medium-128k-instruct, Phi-3-medium-4k-instruct, Phi-3-mini-128k-instruct, Phi-3-mini-4k-instruct, Phi-3-small-128k-instruct, Phi-3-small-8k-instruct
 model_name = "openai/gpt-4o-mini"

diff --git a/samples/python/azure_ai_inference/tools.py b/samples/python/azure_ai_inference/tools.py
@@ -26,9 +26,9 @@
 # by modifying the value of `modelName` in the code below. For this code sample
 # you need a model supporting tools. The following compatible models are
 # available in the GitHub Models service:
-# 
+#
 # Cohere: Cohere-command-r, Cohere-command-r-plus
-# Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, Mistral-small
+# Mistral AI: Mistral-large, Mistral-large-2407, Mistral-Nemo, mistral-small-2503
 # Azure OpenAI: gpt-4o-mini, gpt-4o
 model_name = "openai/gpt-4o-mini"
 
@@ -117,4 +117,4 @@ def get_flight_info(origin_city: str, destination_city: str):
                 model=model_name,
             )
 
-            print(f"Model response = {response.choices[0].message.content}")
+            print(f"Model response = {response.choices[0].message.content}")
diff --git a/samples/python/mistralai/README.md b/samples/python/mistralai/README.md
@@ -16,6 +16,6 @@ To run these scripts, open your terminal and run a command like:
 python3 samples/python/mistralai/basic.py
 ```
 
-* [basic.py](basic.py): basic call to the Mistral-small chat completion API
+* [basic.py](basic.py): basic call to the mistral-small-2503 chat completion API
 * [multi_turn.py](multi_turn.py): multi-turn conversation with the chat completion API
 * [streaming.py](streaming.py): generate a response in streaming mode, token by token