Merge pull request #369 from mistralai/doc/v0.1.32

Update docs to v0.1.32

Author: pandora-s-git

changelog/2025-12-09.mdx

@@ -3,5 +3,5 @@ id: 2025-12-09
 date: 2025-12-09
 ---
 
-- We released Devstral 2 (`devstral-2512`) and Devstral Small 2 (`devstral-small-2512`). <Tag model/>
+- We released Devstral 2 (`devstral-2512`) and Devstral Small 2 (`labs-devstral-small-2512`). <Tag model/>
 - We released Mistral Vibe. <Tag other/>

changelog/2025-12-16.mdx

@@ -0,0 +1,6 @@
+---
+id: 2025-12-16
+date: 2025-12-16
+---
+
+- We released Mistral Small Creative (`labs-mistral-small-creative`) as a Labs model. <Tag model/>

public/img/fc_step1.png

@@ -25,7 +25,7 @@ We provide a variety of different **types** of tools agents can laverage.
 |---|---|
 |[Built-In](tools/built-in)| These are available tools out of the box we provide, these can be called at any given point and all execution happens in our internal environment and cloud.|
 |[Functions](tools/function_calling)| Custom local tools, these are usually functions that are defined in your local environment or code and can be called by the agent, execution happens in your local environment.|
-|[MCP Servers](tools/mcp)| Following the Model Context Protocol, we provide an SDK compatible wich allows you to create or leverage tools from created servers that can be called by the agent, execution happens in your local environment and/or cloud provider hosting the MCP server.|
+|[MCP Servers](tools/mcp)| Following the Model Context Protocol, we provide an SDK compatible which allows you to create or leverage tools from created servers that can be called by the agent, execution happens in your local environment and/or cloud provider hosting the MCP server.|
 
 <SectionTab as="h2" variant="secondary" sectionId="built-in">Built-In</SectionTab>
 
@@ -35,4 +35,4 @@ Below is a list of the built-in tools currently available.
 |[Web Search](tools/built-in/websearch)| Allows searching the web for information, this is useful for finding up to date information or information that is not in the agents knowledge.|
 |[Code Interpreter](tools/built-in/code_interpreter)| Can run code and generate plots, this is useful for data analysis, visualization or running code in a sandboxed environment.|
 |[Image Generation](tools/built-in/image_generation)| Generates images, this is useful for creating images based on a prompt or description.|
-|[Document Library](tools/built-in/document_library)| Provides access to read and search through a vast amount of documents, this is useful for finding information in documents or PDFs and perform RAG (Retrieval Augmented Generation) to answer questions based on very specific information.|
+|[Document Library](tools/built-in/document_library)| Provides access to read and search through a vast amount of documents, this is useful for finding information in documents or PDFs and perform RAG (Retrieval Augmented Generation) to answer questions based on very specific information.|

public/img/fc_step2.png

@@ -30,6 +30,9 @@ As a JSONL file, each line represents a request to the API Endpoint and Model.
 
 The body request will follow the same format as the endpoint you want to run your batching, except the model id that will be provided only during the job creation to start the batch run. Below we provide an example of row with chat completions:
 
+<Tabs>
+    <TabItem value="file" label="Sample Example" default>
+
 ```py
 {
     "custom_id": "2", # An ID as metadata that will be returned in the output file to identify the request
@@ -46,4 +49,43 @@ The body request will follow the same format as the endpoint you want to run you
 }
 ```
 
-For more information regarding completions, visit the [Chat Completions](completion) docs and the corresponding [API Spec](../api/endpoint/chat#operation-chat_completion_v1_chat_completions_post).
+    </TabItem>
+    <TabItem value="output" label="Sample Result">
+
+```py
+{
+  "id": "batch-cd0da3ab-3-a7015699-b9f8-4bdf-928d-c4a29fc0c306", # The ID of the request
+  "custom_id": "2", # The custom ID metadata provided in the input file
+  "response": { # The response body of the request
+    "status_code": 200,
+    "body": {
+      "id": "505471545c4245eeb82dce5409eff9e8",
+      "object": "chat.completion",
+      "model": "ministral-3b-latest",
+      "usage": {
+        "prompt_tokens": 11,
+        "completion_tokens": 256,
+        "total_tokens": 267
+      },
+      "created": 1761577934,
+      "choices": [
+        {
+          "index": 0,
+          "finish_reason": "length",
+          "message": {
+            "role": "assistant",
+            "content": "Choosing the \"best\" French pastry can be quite subjective as it depends on personal preferences. However, here are a few that are often considered among the best:\n\n1. **Croissant**: A buttery, flaky pastry with a crispy exterior and soft, doughy interior. It's often enjoyed for breakfast or as a snack.\n\n2. **Macaron**: A meringue-based confection with a ganache or buttercream filling. They come in various flavors and are known for their delicate texture and vibrant colors.\n\n3. **Éclairs**: Long, thin pastries filled with cream and topped with chocolate icing. They are often enjoyed as a dessert or snack.\n\n4. **Tarte Tatin**: A caramelized apple tart, which is often served upside down with a scoop of ice cream.\n\n5. **Pain au Chocolat**: Similar to a croissant, but filled with chocolate. It's a popular breakfast pastry.\n\n6. **Madeleine**: Small, sponge-like cakes with a distinctive shell-like shape. They are often filled with jam or cream.\n\n7. **Chouquette**: A small, round pastry filled with cream and topped with sugar. It's a popular breakfast pastry.\n\n8.",
+            "tool_calls": null
+          }
+        }
+      ]
+    }
+  },
+  "error": null
+}
+```
+
+    </TabItem>
+</Tabs>
+
+For more information regarding completions, visit the [Chat Completions](completion) docs and the corresponding [API Spec](../api/endpoint/chat#operation-chat_completion_v1_chat_completions_post).