Langchain Model Params
\\
In the previous section, we learned the basic usage of [init_chat_model()](https://example.com/langchain/langchain-init_chat_mode.html).\\
\\
This section will introduce in detail the common control parameters when calling the model, helping you better control the model's behavior.\\
\\
* * *\\
\\
## temperatureββControlling Creativity and Determinism\\
\\
temperature is the most commonly used parameter, with a value range from 0 to 2. It controls the randomness of the model's output.\\
\\
## Example\\
\\
from langchain.chat_models import init_chat_model\\
\\
# Same question, comparison of different temperature\\
\\
question ="Introduce the TUTORIAL platform in one sentence"\\
\\
# temperature=0: output is very deterministic, almost the same every time\\
\\
model_low = init_chat_model("deepseek:deepseek-v4-flash", temperature=0)\\
\\
resp1 = model_low.invoke(question)\\
\\
resp2 = model_low.invoke(question)\\
\\
print(f"temperature=0 Line1times: {resp1.content}")\\
\\
print(f"temperature=0 Line2times: {resp2.content}")\\
\\
print(f"Two times the result is the same: {resp1.content == resp2.content}")\\
\\
print()\\
\\
# temperature=1.5: diverse output, may differ each time\\
\\
model_high = init_chat_model("deepseek:deepseek-v4-flash", temperature=1.5)\\
\\
resp1 = model_high.invoke(question)\\
\\
resp2 = model_high.invoke(question)\\
\\
print(f"temperature=1.5 Line1times: {resp1.content}")\\
\\
print(f"temperature=1.5 Line2times: {resp2.content}")\\
\\
Run result:\\
\\
temperature=0 Line1times: TUTORIAL is a learning platform that provides free tutorials for programming beginners. temperature=0 Line2times: Tutorial (TUTORIAL) is a learning platform that provides free tutorials for programming beginners.Two times the result is the same: True temperature=1.5 Line1times: TUTORIAL is a simple and easy-to-understand introductory learning platform for programming and network technology. temperature=1.5 Line2times: TUTORIALοΌis a Chinese free online technical learning platform beloved by programming novices.\\
\\
| temperature Value | Effect | Applicable Scenario |\\
| --- | --- | --- |\\
| 0 ~ 0.3 | Stable, deterministic output, almost consistent every time | Data extraction, classification, code generation, translation |\\
| 0.5 ~ 0.7 | Moderate creativity, natural output but on topic | Daily conversation, content summarization |\\
| 0.8 ~ 1.2 | Diverse output, more room for creativity | Creative writing, brainstorming |\\
| 1.3 ~ 2.0 | Very random output, may produce unexpected content | Exploratory generation (not recommended for production) |\\
\\
> temperature=0 does not equal "completely identical". Due to floating-point precision differences in the model's internal calculations, there may still be tiny differences in extreme cases. If absolute determinism is needed, some models provide a seed parameter.\\
\\
* * *\\
\\
## max_tokensββControlling Output Length and Cost\\
\\
max_tokens limits the maximum number of Tokens the model can output. One Token is approximately equivalent to 0.75 English words or 0.5 Chinese characters.\\
\\
## Example\\
\\
from langchain.chat_models import init_chat_model\\
\\
model = init_chat_model("deepseek:deepseek-v4-flash", temperature=0)\\
\\
# max_tokens=30: limit output to within 30 Tokens\\
\\
response_short = model.invoke(\\
\\
"Introduce the TUTORIAL platform in detail",\\
\\
max_tokens=30\\
\\
)\\
\\
print(f"Limit 30 tokens ({len(response_short.content)} Character):")\\
\\
print(response_short.content)\\
\\
print()\\
\\
# max_tokens=200: allow longer output\\
\\
response_long = model.invoke(\\
\\
"Introduce the TUTORIAL platform in detail",\\
\\
max_tokens=200\\
\\
)\\
\\
print(f"Limit 200 tokens ({len(response_long.content)} Character):")\\
\\
print(response_long.content)\\
\\
Run result:\\
\\
Limit 30 tokens (42 Character):TUTORIAL is an online learning platform for programming beginners, providing tutorials and examples in various programming languages. Limit 200 tokens (306 Character):TUTORIAL is an online learning platform for programming beginners, providing rich programming tutorials and reference materials. The platform covers HTML, CSS, JavaScript, Python, Java, C, C++, PHP, SQL and other programming languages and technologies. Features of TUTORIAL include:1. Systematic content: from basics to advanced, helping learners gradually master programming knowledge2. Rich examples: each knowledge point comes with runnable code examples3. ...\\
\\
> max_tokens is a hard limit on output length. If set too
YouTip