Mistral Nemo

Description

A 12-billion parameter model featuring a 128k token context window, developed by Mistral in partnership with NVIDIA. The model supports multiple languages including English, French, German, Spanish, Italian, Portuguese, Chinese, Japanese, Korean, Arabic, and Hindi. It includes function calling capabilities and operates under the Apache 2.0 license.

Technical Specifications

Specification	Value
Context Window	128,000 tokens
Parameter Count	12 billion
Context Length	131,072 tokens (128k)
Max Completion Tokens	16,384
Input Modalities	Text
Output Modalities	Text
Quantization	FP8

Pricing

Type	Cost per Million Tokens
Input Tokens	$0.02
Output Tokens	$0.04

Supported Parameters

Parameter	Description
Maximum Tokens	Maximum number of tokens to generate
Temperature	Controls randomness in output
Top P	Nucleus sampling parameter
Top K	Top-k sampling parameter
Stop Sequences	Sequences that stop generation
Frequency Penalty	Reduces repetition of frequent tokens
Presence Penalty	Reduces repetition of any tokens already used
Repetition Penalty	General repetition penalty
Seed	Random seed for reproducibility
Min P	Minimum probability threshold
Response Format	Output format specification

Model Identification

Property	Value
Model Name	Mistral: Mistral Nemo
Model ID	`mistralai/mistral-nemo`
Creator	Mistral AI (in collaboration with NVIDIA)
Release Date	July 19, 2024
License	Apache 2.0

Provider Information

Property	Value
Primary Provider	DeepInfra
Provider Display	DeepInfra \| mistralai/mistral-nemo
Model Variant	Standard

Supported Languages

English
French
German
Spanish
Italian
Portuguese
Chinese
Japanese
Korean
Arabic
Hindi

Capabilities and Features

Function Calling: Supported
Tool Choice Options:
- Literal: none, auto, required
- Type: function
Multipart Input: Supported
Abortable Requests: Supported
Multilingual: 11 languages supported

Usage Example

curl https://api.langmart.ai/v1/chat/completions \
  -H "Authorization: Bearer $LANGMART_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "mistralai/mistral-nemo",
    "messages": [
      {"role": "user", "content": "Hello, how are you?"}
    ]
  }'

Notes

Developed in collaboration between Mistral AI and NVIDIA
Offers a good balance between model size (12B parameters) and context length (128k tokens)
Cost-effective pricing at $0.02/$0.04 per 1M tokens for input/output
Open-source under Apache 2.0 license

Resources

Model Weights: Available via Hugging Face
Chat Interface: Available on LangMart
Model Comparison: Available on LangMart

Data sourced from LangMart.ai on December 23, 2025