Add the SmolLM3 model

.typos.toml

@@ -13,6 +13,7 @@ extend-ignore-identifiers-re = [
     "tese",
     "seperable",
     "Seperable",
+    "setp",
 ]
 
 [files]

README.md

@@ -67,6 +67,22 @@ Please submit requests for new models [here](https://github.com/EricLBuehler/mis
 ## Quick examples
 
 *After following installation instructions*
+
+- 🤗🤗🤗 Run the **SmolLM 3** long-context hybrid-reasoning model with full tool-calling support: [documentation](docs/SMOLLM3.md)  
+  <details>
+    <summary>Show command</summary>
+
+    **Default, easiest:**
+    ```bash
+    ./mistralrs-server -i --isq 8 run -m HuggingFaceTB/SmolLM3-3B
+    ```
+
+    **UQFF prequantized:**
+    ```bash
+    ./mistralrs-server -i run -m EricB/SmolLM3-3B-UQFF -f smollm33b-q4k-0.uqff
+    ```
+  </details>
+
 - 🔊 Run the **Dia 1.6b** model for highly-realistic dialogue generation: [documentation](docs/DIA.md)  
   <details>
     <summary>Show command</summary>
@@ -493,6 +509,7 @@ If you do not specify the architecture, an attempt will be made to use the model
 - `deepseekv3`
 - `qwen3`
 - `qwen3moe`
+- `smollm3`
 
 </details>
 
@@ -576,6 +593,7 @@ Please submit more benchmarks via raising an issue!
 |Mistral 3| | |✅|
 |Llama 4| | |✅|
 |Qwen 3|✅| |✅|
+|SmolLM3| | |✅|
 |Dia 1.6b| | |✅|
 </details>
 
@@ -622,6 +640,7 @@ Please submit more benchmarks via raising an issue!
 |Mistral 3| | | |
 |Llama 4| | | |
 |Qwen 3| | | |
+|SmolLM3|✅| | |
 </details>
 
 <details>
@@ -655,6 +674,7 @@ Please submit more benchmarks via raising an issue!
 |Mistral 3|✅|
 |Llama 4| |
 |Qwen 3| |
+|SmolLM3|✅|
 </details>
 
 ### Using derivative and adapter models

chat_templates/smollm3.jinja

@@ -0,0 +1,89 @@
+{# ───── defaults ───── #}
+{%- if enable_thinking is not defined -%}
+{%- set enable_thinking = true -%}
+{%- endif -%}
+
+{# ───── reasoning mode ───── #}
+{%- if enable_thinking -%}
+  {%- set reasoning_mode = "/think" -%}
+{%- else -%}
+  {%- set reasoning_mode = "/no_think" -%}
+{%- endif -%}
+
+{# ───── header (system message) ───── #}
+{{- "<|im_start|>system\n" -}}
+
+{%- if messages[0].role == "system" -%}
+  {%- set system_message = messages[0].content -%}
+  {%- if "/no_think" in system_message -%}
+    {%- set reasoning_mode = "/no_think" -%}
+  {%- elif "/think" in system_message -%}
+    {%- set reasoning_mode = "/think" -%}
+  {%- endif -%}
+  {%- set custom_instructions = system_message.replace("/no_think", "").replace("/think", "").rstrip() -%}
+{%- endif -%}
+
+{%- if "/system_override" in system_message -%}
+  {{- custom_instructions.replace("/system_override", "").rstrip() -}}
+  {{- "<|im_end|>\n" -}}
+{%- else -%}
+  {{- "## Metadata\n\n" -}}
+  {{- "Knowledge Cutoff Date: June 2025\n" -}}
+  {%- set today = strftime_now("%d %B %Y") -%}
+  {{- "Today Date: " ~ today ~ "\n" -}}
+  {{- "Reasoning Mode: " + reasoning_mode + "\n\n" -}}
+
+  {{- "## Custom Instructions\n\n" -}}
+  {%- if custom_instructions -%}
+    {{- custom_instructions + "\n\n" -}}
+  {%- elif reasoning_mode == "/think" -%}
+    {{- "You are a helpful AI assistant named SmolLM, trained by Hugging Face. Your role as an assistant involves thoroughly exploring questions through a systematic thinking process before providing the final precise and accurate solutions. This requires engaging in a comprehensive cycle of analysis, summarizing, exploration, reassessment, reflection, backtracking, and iteration to develop well-considered thinking process. Please structure your response into two main sections: Thought and Solution using the specified format: <think> Thought section </think> Solution section. In the Thought section, detail your reasoning process in steps. Each step should include detailed considerations such as analysing questions, summarizing relevant findings, brainstorming new ideas, verifying the accuracy of the current steps, refining any errors, and revisiting previous steps. In the Solution section, based on various attempts, explorations, and reflections from the Thought section, systematically present the final solution that you deem correct. The Solution section should be logical, accurate, and concise and detail necessary steps needed to reach the conclusion.\n\n" -}}
+  {%- else -%}
+    {{- "You are a helpful AI assistant named SmolLM, trained by Hugging Face.\n\n" -}}
+  {%- endif -%}
+
+  {%- if xml_tools or python_tools -%}
+    {{- "### Tools\n\n" -}}
+    {%- if xml_tools -%}
+      {%- set ns = namespace(xml_tool_string="You may call one or more functions to assist with the user query.\nYou are provided with function signatures within <tools></tools> XML tags:\n\n<tools>\n") -%}
+      {%- for tool in xml_tools[:] -%} {# The slicing makes sure that xml_tools is a list #}
+        {%- set ns.xml_tool_string = ns.xml_tool_string ~ (tool | string) ~ "\n" -%}
+      {%- endfor -%}
+      {%- set xml_tool_string = ns.xml_tool_string + "</tools>\n\nFor each function call, return a json object with function name and arguments within <tool_call></tool_call> XML tags:\n<tool_call>\n{\"name\": <function-name>, \"arguments\": <args-json-object>}\n</tool_call>" -%}
+      {{- xml_tool_string -}}
+    {%- endif -%}
+    {%- if python_tools -%}
+      {%- set ns = namespace(python_tool_string="When you send a message containing Python code between '<code>' and '</code>' tags, it will be executed in a stateful Jupyter notebook environment, and you will then be given the output to continued reasoning in an agentic loop.\n\nYou can use the following tools in your python code like regular functions:\n<tools>\n") -%}
+      {%- for tool in python_tools[:] -%} {# The slicing makes sure that python_tools is a list #}
+        {%- set ns.python_tool_string = ns.python_tool_string ~ (tool | string) ~ "\n" -%}
+      {%- endfor -%}
+      {%- set python_tool_string = ns.python_tool_string + "</tools>\n\nThe state persists between code executions: so variables that you define in one step are still available thereafter." -%}
+      {{- python_tool_string -}}
+    {%- endif -%}
+    {{- "\n\n" -}}
+    {{- "<|im_end|>\n" -}}
+  {%- endif -%}
+{%- endif -%}
+{# ───── main loop ───── #}
+{%- for message in messages -%}
+    {%- set content = message.content if message.content is string else "" -%}
+    {%- if message.role == "user" -%}
+        {{ "<|im_start|>" + message.role + "\n"  + content + "<|im_end|>\n" }}
+    {%- elif message.role == "assistant" -%}
+        {%- if reasoning_mode == "/think" -%}
+            {{ "<|im_start|>assistant\n" + content.lstrip("\n") + "<|im_end|>\n" }}
+        {%- else -%}
+            {{ "<|im_start|>assistant\n" + "<think>\n\n</think>\n" + content.lstrip("\n") + "<|im_end|>\n" }}
+        {%- endif -%}
+    {%- elif message.role == "tool" -%}
+    {{ "<|im_start|>" + "user\n"  + content + "<|im_end|>\n" }}
+    {%- endif -%}
+{%- endfor -%}
+{# ───── generation prompt ───── #}
+{%- if add_generation_prompt -%}
+    {%- if reasoning_mode == "/think" -%}
+        {{ "<|im_start|>assistant\n" }}
+    {%- else -%}
+        {{ "<|im_start|>assistant\n" + "<think>\n\n</think>\n"  }}
+    {%- endif -%}
+{%- endif -%}

docs/ANYMOE.md

@@ -126,7 +126,7 @@ runner = Runner(
 
 res = runner.send_chat_completion_request(
     ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {"role": "user", "content": "Tell me a story about the Rust type system."}
         ],

docs/DEEPSEEKV2.md

@@ -35,7 +35,7 @@ while True:
     prompt = input(">>> ")
     messages.append({"role": "user", "content": prompt})
     completion = client.chat.completions.create(
-        model="ignore",
+        model="default",
         messages=messages,
         max_tokens=256,
         frequency_penalty=1.0,
@@ -60,7 +60,7 @@ runner = Runner(
 
 res = runner.send_chat_completion_request(
     ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {"role": "user", "content": "Tell me a story about the Rust type system."}
         ],

docs/DEEPSEEKV3.md

@@ -44,7 +44,7 @@ while True:
     prompt = input(">>> ")
     messages.append({"role": "user", "content": prompt})
     completion = client.chat.completions.create(
-        model="ignore",
+        model="default",
         messages=messages,
         max_tokens=256,
         frequency_penalty=1.0,
@@ -69,7 +69,7 @@ runner = Runner(
 
 res = runner.send_chat_completion_request(
     ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {"role": "user", "content": "Tell me a story about the Rust type system."}
         ],

docs/DIA.md

@@ -29,7 +29,7 @@ client = OpenAI(api_key="foobar", base_url="http://localhost:1234/v1/")
 text_to_speak = "[S1] mistral r s is a local LLM inference engine. [S2] You can run text and vision models, and also image generation and speech generation. [S1] There is agentic web search, tool calling, and a convenient Python API. [S2] Check it out on github."
 
 response = client.audio.speech.create(
-    model="ignore", voice="N/A", input=text_to_speak, response_format="wav"
+    model="default", voice="N/A", input=text_to_speak, response_format="wav"
 )
 
 output_path = Path("output.wav")

docs/FLUX.md

@@ -30,7 +30,7 @@ from openai import OpenAI
 client = OpenAI(api_key="foobar", base_url="http://localhost:1234/v1/")
 
 result = client.images.generate(
-    model="ignore",
+    model="default",
     prompt="A vibrant sunset in the mountains, 4k, high quality.",
     n=1,
 )

docs/GEMMA2.md

@@ -19,7 +19,7 @@ while True:
     prompt = input(">>> ")
     messages.append({"role": "user", "content": prompt})
     completion = client.chat.completions.create(
-        model="ignore",
+        model="default",
         messages=messages,
         max_tokens=256,
         frequency_penalty=1.0,
@@ -44,7 +44,7 @@ runner = Runner(
 
 res = runner.send_chat_completion_request(
     ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {"role": "user", "content": "Tell me a story about the Rust type system."}
         ],

docs/GEMMA3.md

@@ -63,7 +63,7 @@ client = OpenAI(api_key="foobar", base_url="http://localhost:1234/v1/")
 
 
 completion = client.chat.completions.create(
-    model="ignore",
+    model="default",
     messages=[
         {
             "role": "user",
@@ -160,7 +160,7 @@ runner = Runner(
 
 res = runner.send_chat_completion_request(
     ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {
                 "role": "user",

docs/GLM4.md

@@ -19,7 +19,7 @@ while True:
     prompt = input(">>> ")
     messages.append({"role": "user", "content": prompt})
     completion = client.chat.completions.create(
-        model="ignore",
+        model="default",
         messages=messages,
         max_tokens=256,
         frequency_penalty=1.0,
@@ -44,7 +44,7 @@ runner = Runner(
 
 res = runner.send_chat_completion_request(
     ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {"role": "user", "content": "Tell me a story about the Rust type system."}
         ],

docs/HTTP.md

@@ -46,7 +46,7 @@ client = openai.OpenAI(
 )
 
 completion = client.chat.completions.create(
-model="ignore",
+model="default",
 messages=[
     {"role": "system", "content": "You are Mistral.rs, an AI assistant."},
     {"role": "user", "content": "Write a story about Rust error handling."}
@@ -62,7 +62,7 @@ curl http://localhost:8080/v1/chat/completions \
 -H "Content-Type: application/json" \
 -H "Authorization: Bearer EMPTY" \
 -d '{
-"model": "ignore",
+"model": "default",
 "messages": [
 {
     "role": "system",
@@ -116,7 +116,7 @@ client = openai.OpenAI(
 )
 
 completion = client.completions.create(
-    model="ignore",
+    model="default",
     prompt="What is Rust?",
     max_tokens=256,
     frequency_penalty=1.0,
@@ -133,7 +133,7 @@ curl http://localhost:8080/v1/completions \
 -H "Content-Type: application/json" \
 -H "Authorization: Bearer EMPTY" \
 -d '{
-"model": "ignore",
+"model": "default",
 "prompt": "What is Rust?"
 }'
 ```

docs/IDEFICS2.md

@@ -58,7 +58,7 @@ from openai import OpenAI
 client = OpenAI(api_key="foobar", base_url="http://localhost:1234/v1/")
 
 completion = client.chat.completions.create(
-    model="ignore",
+    model="default",
     messages=[
         {
             "role": "user",
@@ -155,7 +155,7 @@ runner = Runner(
 
 res = runner.send_chat_completion_request(
     ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {
                 "role": "user",

docs/IDEFICS3.md

@@ -120,7 +120,7 @@ from openai import OpenAI
 client = OpenAI(api_key="foobar", base_url="http://localhost:1234/v1/")
 
 completion = client.chat.completions.create(
-    model="ignore",
+    model="default",
     messages=[
         {
             "role": "user",
@@ -217,7 +217,7 @@ runner = Runner(
 
 res = runner.send_chat_completion_request(
     ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {
                 "role": "user",

docs/LLAMA4.md

@@ -82,7 +82,7 @@ client = OpenAI(api_key="foobar", base_url="http://localhost:1234/v1/")
 
 
 completion = client.chat.completions.create(
-    model="ignore",
+    model="default",
     messages=[
         {
             "role": "user",
@@ -181,7 +181,7 @@ runner = Runner(
 
 res = runner.send_chat_completion_request(
     ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {
                 "role": "user",

docs/LLaVA.md

@@ -68,7 +68,7 @@ from openai import OpenAI
 client = OpenAI(api_key="foobar", base_url="http://localhost:1234/v1/")
 
 completion = client.chat.completions.create(
-    model="ignore",
+    model="default",
     messages=[
         {
             "role": "user",
@@ -164,7 +164,7 @@ runner = Runner(
 
 res = runner.send_chat_completion_request(
     ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {
                 "role": "user",

docs/MCP/README.md

@@ -150,7 +150,7 @@ runner = mistralrs.Runner(
 # Use the model - tools are automatically available
 res = runner.send_chat_completion_request(
     mistralrs.ChatCompletionRequest(
-        model="ignore",
+        model="default",
         messages=[
             {"role": "user", "content": "List the files in the current directory and create a test.txt file"}
         ],