Mindket

all-mpnet-base-v2

Newversion 1 · updated

Overview

Turns sentences and paragraphs into 768-number vectors so you can search, cluster and compare text by meaning. It suits developers and data teams working in English.

all-mpnet-base-v2 is a kit in the Developer category, listed by Mindket. It is free. This is version 1, last updated on . Download it on this page, a zip of its files. Nothing in it runs by itself.

At a glance

You need
  • Python on your computer
  • The sentence-transformers package
  • About 438 MB of disk for the F32 weights, or about 55 MB at 4-bit by Hugging Face's estimate
Runs in
  • Python with the sentence-transformers library
  • Python with Hugging Face Transformers, where you add the pooling step yourself
  • Text Embeddings Inference, run in Docker on a CPU or an NVIDIA GPU
  • Windows, macOS or Linux with Python
Start with
pip install -U sentence-transformers

Platform

DockerContainer
WindowsOperating system
macOSOperating system
LinuxOperating system

About

What it is

all-mpnet-base-v2 is a sentence-transformers model that converts sentences and paragraphs into 768-number vectors. Texts with similar meaning end up close together, which makes the vectors useful for semantic search and clustering. It starts from the MPNet base model and was fine-tuned on a dataset of 1B sentence pairs.

Who it suits

It suits developers, data teams and researchers who need to compare English text by meaning instead of by exact words.

IndustryWho uses itHow they use itExample
SoftwareSearch engineersTurn documents and queries into vectors and rank by closenessA team embeds help articles and finds the ones nearest a user's question
SupportSupport teamsMatch a new message to similar past onesA team groups incoming tickets that describe the same problem
ResearchNLP researchersCluster texts by meaningA researcher groups survey answers into themes

If your text is not in English, you can skip it, since the record lists English only.

Where to use it

  • Python with the sentence-transformers library
  • Python with Hugging Face Transformers, where you add the pooling step yourself
  • Text Embeddings Inference, run in Docker on a CPU or an NVIDIA GPU
  • Windows, macOS or Linux with Python
Use caseWhoWhat they get back
Semantic searchDeveloperVectors you can compare to find the closest text
ClusteringData analystVectors you can group into topics
Similar text matchingSupport or research teamA closeness score between two pieces of text

How to set it up

  • Python on your computer
  • The sentence-transformers package
  • About 438 MB of disk for the F32 weights, or about 55 MB at 4-bit by Hugging Face's estimate
  1. Install sentence-transformers with pip.
  2. Run the Python sample, which loads the model by name from the Hub.
  3. Optionally download the files with the Hugging Face tools if you want them on disk first.
pip install -U sentence-transformers
pip install -U huggingface_hub
hf download sentence-transformers/all-mpnet-base-v2

How to use it

You load the model, pass it a list of sentences, and call encode to get one vector per sentence. The card also shows a route through plain Transformers, where you apply mean pooling and normalize the vectors yourself.

from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]

model = SentenceTransformer('sentence-transformers/all-mpnet-base-v2')
embeddings = model.encode(sentences)
print(embeddings)
ToolWhat it does
SentenceTransformerLoads the model by its name
model.encodeTurns a list of sentences into vectors
AutoTokenizer and AutoModelThe plain Transformers route, where you pool and normalize
Text Embeddings InferenceServes the model from a container, with an OpenAI style embeddings route

Example

You encode two short sentences, one saying this is an example and one saying each sentence is converted. The script prints two vectors of 768 numbers each, one per sentence, which you can compare for closeness.

For newcomers and experts

An embedding is a list of numbers that stands for the meaning of a text. Sentences that mean similar things get lists that sit close together. A first small thing to try is the sample above with two sentences of your own.

For experts, the model was trained with a contrastive objective. Given one sentence from a pair, it must pick the true partner out of a set of randomly sampled sentences. The team ran it on 7 TPUs v3-8 during a Hugging Face community week using JAX and Flax. With plain Transformers you take the mean of token embeddings, weighted by the attention mask, then normalize.

  1. Your sentences go in
  2. The tokenizer splits them into tokens
  3. The MPNet model builds token embeddings
  4. Mean pooling combines them per sentence
  5. You get one 768-number vector for each

Good to know

  • Size: 109M parameters and 768 output dimensions
  • Language: English
  • License: Apache 2.0
  • Cost to run: none named
  • Data it reaches: the sample loads the model from the Hub by name, and the source names no other access
  • Not for: text in other languages, since the record lists English only

README and files

README

all-mpnet-base-v2

Turns sentences and paragraphs into 768-number vectors so you can search, cluster and compare text by meaning. It suits developers and data teams working in English.

Made by sentence-transformers on Hugging Face. These notes are Mindket's, written from its model page and model card; the model's own files and license are on Hugging Face.

About

What it is: all-mpnet-base-v2 is a sentence-transformers model that converts sentences and paragraphs into 768-number vectors. Texts with similar meaning end up close together, which makes the vectors useful for semantic search and clustering. It starts from the MPNet base model and was fine-tuned on a dataset of 1B sentence pairs.

Who it suits: It suits developers, data teams and researchers who need to compare English text by meaning instead of by exact words.

Industry | Who uses it | How they use it | Example
Software | Search engineers | Turn documents and queries into vectors and rank by closeness | A team embeds help articles and finds the ones nearest a user's question
Support | Support teams | Match a new message to similar past ones | A team groups incoming tickets that describe the same problem
Research | NLP researchers | Cluster texts by meaning | A researcher groups survey answers into themes

If your text is not in English, you can skip it, since the record lists English only.

Where to use it:

  • Python with the sentence-transformers library
  • Python with Hugging Face Transformers, where you add the pooling step yourself
  • Text Embeddings Inference, run in Docker on a CPU or an NVIDIA GPU
  • Windows, macOS or Linux with Python
Use case | Who | What they get back
Semantic search | Developer | Vectors you can compare to find the closest text
Clustering | Data analyst | Vectors you can group into topics
Similar text matching | Support or research team | A closeness score between two pieces of text

How to set it up:

  • Python on your computer
  • The sentence-transformers package
  • About 438 MB of disk for the F32 weights, or about 55 MB at 4-bit by Hugging Face's estimate
  1. Install sentence-transformers with pip.
  2. Run the Python sample, which loads the model by name from the Hub.
  3. Optionally download the files with the Hugging Face tools if you want them on disk first.
pip install -U sentence-transformers
pip install -U huggingface_hub
hf download sentence-transformers/all-mpnet-base-v2

How to use it: You load the model, pass it a list of sentences, and call encode to get one vector per sentence. The card also shows a route through plain Transformers, where you apply mean pooling and normalize the vectors yourself.

from sentence_transformers import SentenceTransformer
sentences = ["This is an example sentence", "Each sentence is converted"]

model = SentenceTransformer('sentence-transformers/all-mpnet-base-v2')
embeddings = model.encode(sentences)
print(embeddings)
Tool | What it does
SentenceTransformer | Loads the model by its name
model.encode | Turns a list of sentences into vectors
AutoTokenizer and AutoModel | The plain Transformers route, where you pool and normalize
Text Embeddings Inference | Serves the model from a container, with an OpenAI style embeddings route

Example: You encode two short sentences, one saying this is an example and one saying each sentence is converted. The script prints two vectors of 768 numbers each, one per sentence, which you can compare for closeness.

For newcomers and experts: An embedding is a list of numbers that stands for the meaning of a text. Sentences that mean similar things get lists that sit close together. A first small thing to try is the sample above with two sentences of your own.

For experts, the model was trained with a contrastive objective. Given one sentence from a pair, it must pick the true partner out of a set of randomly sampled sentences. The team ran it on 7 TPUs v3-8 during a Hugging Face community week using JAX and Flax. With plain Transformers you take the mean of token embeddings, weighted by the attention mask, then normalize.

Your sentences go in
The tokenizer splits them into tokens
The MPNet model builds token embeddings
Mean pooling combines them per sentence
You get one 768-number vector for each

Good to know:

  • Size: 109M parameters and 768 output dimensions
  • Language: English
  • License: Apache 2.0
  • Cost to run: none named
  • Data it reaches: the sample loads the model from the Hub by name, and the source names no other access
  • Not for: text in other languages, since the record lists English only

On Hugging Face, 16 November 2026

  • Task: sentence similarity
  • Library: sentence-transformers
  • Parameters: 109M
  • Weights: about 438 MB in F32; about 55 MB at 4-bit (estimate)
  • Formats: Safetensors
  • Downloads in the last 30 days: 19,940,706
  • Likes: 1,406
  • License: apache-2.0
  • Languages: en
  • Quantized versions: 8
  • Fine-tunes: 416
  • Tags: sentence-transformers, pytorch, onnx, safetensors, openvino, mpnet, fill-mask, feature-extraction, sentence-similarity, transformers, text-embeddings-inference

Links

Get the model

Download its files with Hugging Face's command line:

pip install -U huggingface_hub
hf download sentence-transformers/all-mpnet-base-v2

Then follow the usage in its model card. Nothing in this kit runs by itself.

Files
README.md  5 KB

1 file, 5 KB, downloaded as a zip.

What’s new

  • Version 1 · 2026-11-16

Compare with similar

Listed Hugging Face models in the same category that share the most tags with this one.

ProductDownloads, last 30 daysLikesParametersLicenseLast change
all-mpnet-base-v2 This page19,940,7061,406109Mapache-2.0
all-MiniLM-L6-v2Shares openvino · sentence-similarity · text-embeddings-inference174,740,1846,10422.7Mapache-2.0
paraphrase-multilingual-MiniLM-L12-v2Shares openvino · sentence-similarity · text-embeddings-inference26,540,1251,505118Mapache-2.0
multilingual-e5-largeShares openvino · sentence-similarity · text-embeddings-inference7,120,5551,284560Mmit
multilingual-e5-baseShares openvino · sentence-similarity · text-embeddings-inference4,750,367422278Mmit
bge-small-en-v1.5Shares sentence-similarity · text-embeddings-inference · sentence-transformers45,750,54867133.4Mmit
Numbers as Mindket last read them, from 2026-11-12 to 2026-11-16.

Reviews

No ratings yet

Write a review

No reviews yet.

What people say

Posts and comments that name this project on Hacker News, Bluesky, X, Reddit, Facebook and YouTube, found by a weekly search and on Trendshift. Mindket doesn’t write or check them. Open one to read it where it was posted.

Searched

More from Mindket

See all from Mindket

multilingual-e5-base

intfloat · Kit

Turns text in about 100 languages into vectors so you can compare meaning across them, for search and matching. Made for developers who build multilingual retr…

New

sprite-gen

aldegad · Skill

Turns one base drawing of a character into game-ready sprites, either an atlas with a manifest or transparent motion loops, for game developers working in Code…

New4097

Monocolor Editorial Print

yanliudesign · Skill

Makes original one-ink and two-ink editorial images from a theme, an object or your photo, for posters, zines, portraits and packaging. Returns the image, the…

New3662

Similar in Developer

See all in Developer

VibeSkills

foryourhealth111-pixel · Skill

Picks the right skills from a large installed pool and breaks each task into checkable steps for your coding agent. In its own benchmark, mean reward rose 21.1…

New3821

claude-bughunter

elementalsouls · Skill

Gives Claude Code 83 skills and 15 slash commands for bug bounty hunting and authorized red-team work, from web vulnerability classes to enterprise identity at…

New6853

Runs Qwen3.8-27B with a vocabulary cut down to ASCII, so a single 16 GB graphics card can hold a much longer prompt. A GGUF build for English text and code wit…

New