chore(docs): auto-update llms.txt & llms-full.txt

Author: pandora-s-git

static/llms-full.txt

@@ -3117,7 +3117,7 @@ Now that we have our document library agent ready, we can search them on demand
 
 ```py
 response = client.beta.conversations.start(
-    agent_id=image_agent.id,
+    agent_id=library_agent.id,
     inputs="How does the vision encoder for pixtral 12b work"
 )
 ```
@@ -11711,11 +11711,11 @@ Currently we have two reasoning models:
 - `magistral-medium-latest`: Our more powerful reasoning model balancing performance and cost.
 
 :::info
-Currently, `-latest` points to `-2507`, our most recent version of our reasoning models. If you were previously using `-2506`, a **migration** regarding the thinking chunks is required.
-- `-2507` **(new)**: Uses tokenized thinking chunks via control tokens, providing the thinking traces in different types of content chunks.
+Currently, `-latest` points to `-2509`, our most recent version of our reasoning models. If you were previously using `-2506`, a **migration** regarding the thinking chunks is required.
+- `-2507` & `-2509` **(new)**: Uses tokenized thinking chunks via control tokens, providing the thinking traces in different types of content chunks.
 - `-2506` **(old)**: Used `<think>\n` and `\n</think>\n` tags as strings to encapsulate the thinking traces for input and output within the same content type.  
 <Tabs groupId="version">
-  <TabItem value="2507" label="2507 (new)" default>
+  <TabItem value="2509" label="2507/2509 (new)" default>
 ```json
 [
   {
@@ -11794,7 +11794,33 @@ To have the best performance out of our models, we recommend having the followin
 <summary><b>System Prompt</b></summary>
 
 <Tabs groupId="version">
-  <TabItem value="2507" label="2507 (new)" default>
+  <TabItem value="2509" label="2509 (new)" default>
+```json
+{
+  "role": "system",
+  "content": [
+    {
+      "type": "text",
+      "text": "# HOW YOU SHOULD THINK AND ANSWER\n\nFirst draft your thinking process (inner monologue) until you arrive at a response. Format your response using Markdown, and use LaTeX for any mathematical equations. Write both your thoughts and the response in the same language as the input.\n\nYour thinking process must follow the template below:"
+    },
+    {
+      "type": "thinking",
+      "thinking": [
+        {
+          "type": "text",
+          "text": "Your thoughts or/and draft, like working through an exercise on scratch paper. Be as casual and as long as you want until you are confident to generate the response to the user."
+        }
+      ]
+    },
+    {
+      "type": "text",
+      "text": "Here, provide a self-contained response."
+    }
+  ]
+}
+```
+  </TabItem>
+  <TabItem value="2507" label="2507" default>
 ```json
 {
   "role": "system",
@@ -11915,7 +11941,7 @@ curl --location "https://api.mistral.ai/v1/chat/completions" \
 </Tabs>
 
 <Tabs groupId="version">
-  <TabItem value="2507" label="2507 (new)" default>
+  <TabItem value="2509" label="2507/2509 (new)" default>
 The output of the model will include different chunks of content, but mostly a `thinking` type with the reasoning traces and a `text` type with the answer like so:
 ```json
 "content": [
@@ -13101,10 +13127,12 @@ Vision capabilities enable models to analyze images and provide insights based o
 For more specific use cases regarding document parsing and data extraction we recommend taking a look at our Document AI stack [here](../document_ai/document_ai_overview).
 
 ## Models with Vision Capabilities:
+- Mistral Medium 3.1 2508 (`mistral-medium-latest`)
+- Mistral Small 3.2 2506 (`mistral-small-latest`)
+- Magistral Small 1.2 2509 (`magistral-small-latest`)
+- Magistral Medium 1.2 2509 (`magistral-medium-latest`)
 - Pixtral 12B (`pixtral-12b-latest`)
 - Pixtral Large 2411 (`pixtral-large-latest`)
-- Mistral Medium 2505 (`mistral-medium-latest`)
-- Mistral Small 2503 (`mistral-small-latest`)
 
 ## Passing an Image URL
 If the image is hosted online, you can simply provide the URL of the image in the request. This method is straightforward and does not require any encoding.
@@ -16214,14 +16242,16 @@ Mistral provides two types of models: open models and premier models.
 | Model               | Weight availability|Available via API| Description | Max Tokens| API Endpoints|Version|
 |--------------------|:--------------------:|:--------------------:|:--------------------:|:--------------------:|:--------------------:|:--------------------:|
 | Mistral Medium 3.1 | | :heavy_check_mark: | Our frontier-class multimodal model released August 2025. Improving tone and performance. Read more about Medium 3 in our [blog post](https://mistral.ai/news/mistral-medium-3/) | 128k | `mistral-medium-2508` | 25.08|
-| Magistral Medium 1.1 | | :heavy_check_mark: | Our frontier-class reasoning model released July 2025. | 40k | `magistral-medium-2507` | 25.07|
+| Magistral Medium 1.2 | | :heavy_check_mark: | Our frontier-class reasoning model update released September 2025 with vision support. | 128k | `magistral-medium-2509` | 25.09|
 | Codestral 2508 | | :heavy_check_mark: | Our cutting-edge language model for coding released end of July 2025, Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. Learn more in our [blog post](https://mistral.ai/news/codestral-25-08/) | 256k  | `codestral-2508` | 25.08|
 | Voxtral Mini Transcribe | | :heavy_check_mark: | An efficient audio input model, fine-tuned and optimized for transcription purposes only. | | `voxtral-mini-2507` via `audio/transcriptions` | 25.07|
 | Devstral Medium | | :heavy_check_mark: | An enterprise grade text model, that excels at using tools to explore codebases, editing multiple files and power software engineering agents. Learn more in our [blog post](https://mistral.ai/news/devstral-2507) | 128k  | `devstral-medium-2507` | 25.07|
 | Mistral OCR 2505 | | :heavy_check_mark: | Our OCR service powering our Document AI stack that enables our users to extract interleaved text and images |  | `mistral-ocr-2505` | 25.05|
 | Magistral Medium 1 | | :heavy_check_mark: | Our first frontier-class reasoning model released June 2025. Learn more in our [blog post](https://mistral.ai/news/magistral/) | 40k | `magistral-medium-2506` | 25.06|
 | Ministral 3B | | :heavy_check_mark: | World’s best edge model. Learn more in our [blog post](https://mistral.ai/news/ministraux/) | 128k  | `ministral-3b-2410` | 24.10|
 | Ministral 8B | :heavy_check_mark: <br/> [Mistral Research License](https://mistral.ai/licenses/MRL-0.1.md)| :heavy_check_mark: |Powerful edge model with extremely high performance/price ratio. Learn more in our [blog post](https://mistral.ai/news/ministraux/) | 128k  | `ministral-8b-2410` | 24.10|
+| Magistral Medium 1.1 | | :heavy_check_mark: | Our frontier-class reasoning model released July 2025. | 40k | `magistral-medium-2507` | 25.07|
+| Magistral Medium 1.0 | | :heavy_check_mark: | Our frontier-class reasoning model released June 2025. Learn more in our [blog post](https://mistral.ai/news/magistral/) | 40k | `magistral-medium-2506` | 25.06|
 | Mistral Medium 3 | | :heavy_check_mark: | Our frontier-class multimodal model released May 2025. Learn more in our [blog post](https://mistral.ai/news/mistral-medium-3/) | 128k | `mistral-medium-2505` | 25.05|
 | Codestral 2501 | | :heavy_check_mark: | Our cutting-edge language model for coding with the second version released January 2025, Codestral specializes in low-latency, high-frequency tasks such as fill-in-the-middle (FIM), code correction and test generation. Learn more in our [blog post](https://mistral.ai/news/codestral-2501/) | 256k  | `codestral-2501` | 25.01|
 | Mistral Large 2.1 |:heavy_check_mark: <br/> [Mistral Research License](https://mistral.ai/licenses/MRL-0.1.md)| :heavy_check_mark: | Our top-tier large model for high-complexity tasks with the lastest version released November 2024. Learn more in our [blog post](https://mistral.ai/news/pixtral-large/) | 128k  | `mistral-large-2411` | 24.11|
@@ -16235,6 +16265,7 @@ Mistral provides two types of models: open models and premier models.
 
 | Model               | Weight availability|Available via API| Description | Max Tokens| API Endpoints|Version|
 |--------------------|:--------------------:|:--------------------:|:--------------------:|:--------------------:|:--------------------:|:--------------------:|
+| Magistral Small 1.2 | :heavy_check_mark: <br/> Apache2 | :heavy_check_mark: | Our small reasoning model released September 2025 with vision support. | 128k | `magistral-small-2509` | 25.09|
 | Magistral Small 1.1 | :heavy_check_mark: <br/> Apache2 | :heavy_check_mark: | Our small reasoning model released July 2025. | 40k | `magistral-small-2507` | 25.07|
 | Voxtral Small | :heavy_check_mark: <br/> Apache2 | :heavy_check_mark: | Our first model with audio input capabilities for instruct use cases.  | 32k  | `voxtral-small-2507` | 25.07|
 | Voxtral Mini | :heavy_check_mark: <br/> Apache2 | :heavy_check_mark: | A mini version of our first audio input model.  | 32k | `voxtral-mini-2507` | 25.07|
@@ -16255,8 +16286,8 @@ it is recommended to use the dated versions of the Mistral AI API.
 Additionally, be prepared for the deprecation of certain endpoints in the coming months.
 
 Here are the details of the available versions:
-- `magistral-medium-latest`: currently points to `magistral-medium-2507`. 
-- `magistral-small-latest`: currently points to `magistral-small-2507`. 
+- `magistral-medium-latest`: currently points to `magistral-medium-2509`. 
+- `magistral-small-latest`: currently points to `magistral-small-2509`. 
 - `mistral-medium-latest`: currently points to `mistral-medium-2508`. 
 - `mistral-large-latest`: currently points to `mistral-medium-2508`, previously `mistral-large-2411`. 
 - `pixtral-large-latest`: currently points to `pixtral-large-2411`.