Nomic AI website screenshot

Nomic AI

Nomic AI builds open and accessible AI infrastructure. The company is known for the open-source Nomic Embed family (nomic-embed-text-v1.5, nomic-embed-vision-v1.5) of multilingual text and vision embedding models, the Nomic Atlas platform for exploring, labelling, and operationalising unstructured data via interactive 2D maps, and the GPT4All open-source ecosystem for running large language models locally on consumer CPUs and GPUs. The Nomic Atlas REST API at api-atlas.nomic.ai exposes endpoints for datasets, maps, embeddings, file parsing, and task status, with official Python and TypeScript SDKs.

9 APIs 0 Features
EmbeddingsVector DatabaseData ExplorationLLMOpen SourceRAGAtlas

APIs

Nomic Atlas API

REST API for the Nomic Atlas platform. Exposes endpoints for creating and managing datasets, building 2D semantic maps over text and image data, querying and updating points, an...

Nomic Embedding API

Hosted embedding endpoint backed by the Nomic Embed model family (nomic-embed-text-v1.5 and nomic-embed-vision-v1.5). Returns dense multilingual text and image vectors with Matr...

Nomic Atlas File Parsing API

Upload, parse, and extract endpoints for ingesting PDFs and other documents into Atlas as structured text and tables prior to embedding and map building.

Nomic Atlas Task Status API

Polling endpoint that returns the status of long-running Atlas jobs (map builds, embeddings, dataset operations).

Nomic Python SDK

Official Python client (nomic) for the Atlas platform, embedding API, and dataset/map workflows.

Nomic TypeScript SDK

Official TypeScript / JavaScript client for the Nomic Atlas API.

GPT4All

Open-source ecosystem for running large language models locally on consumer hardware. Ships a desktop chat client, a local OpenAI-compatible HTTP server, and Python and TypeScri...

nomic-embed-text-v1.5 (Open Weights)

Open-weights multilingual text embedding model published on Hugging Face, Apache-2.0 licensed. Supports Matryoshka-style truncatable embeddings and task-type prefixes.

nomic-embed-vision-v1.5 (Open Weights)

Open-weights vision embedding model that shares an embedding space with nomic-embed-text-v1.5 to support cross-modal retrieval. Published on Hugging Face.

Collections

Nomic Atlas API

OPEN

Pricing Plans

Nomic Ai Plans Pricing

1 plans

PLANS

Rate Limits

Nomic Ai Rate Limits

2 limits

RATE LIMITS

FinOps

Nomic Ai Finops

FINOPS

Resources

🔗
LinkedIn
LinkedIn
🔗
Website
Website
🔗
Documentation
Documentation
👥
GitHub
GitHub
🔗
HuggingFace
HuggingFace
🔗
Plans
Plans
🔗
RateLimits
RateLimits
🔗
FinOps
FinOps

Sources

Raw ↑ 
opencollection: 1.0.0
info:
  name: Nomic Atlas API
  version: '1.0'
request:
  auth:
    type: bearer
    token: '{{bearerToken}}'
items:
- info:
    name: Embeddings
    type: folder
  items:
  - info:
      name: Generate text embeddings
      type: http
    http:
      method: POST
      url: https://api-atlas.nomic.ai/v1/embedding/text
      body:
        type: json
        data: '{}'
    docs: Generate embeddings for one or more text inputs using a Nomic embedding model.
- info:
    name: Files
    type: folder
  items:
  - info:
      name: Get presigned URLs to upload files
      type: http
    http:
      method: POST
      url: https://api-atlas.nomic.ai/v1/upload
      body:
        type: json
        data: '{}'
    docs: Get URLs to upload files for use in other tasks (parsing, extraction, etc.).
  - info:
      name: Parse a file into structured information
      type: http
    http:
      method: POST
      url: https://api-atlas.nomic.ai/v1/parse
      body:
        type: json
        data: '{}'
    docs: Parse a PDF file with configurable chunking strategies and optional embedding generation. Returns a task id for
      asynchronous polling.
- info:
    name: Tasks
    type: folder
  items:
  - info:
      name: Get the status of a task
      type: http
    http:
      method: GET
      url: https://api-atlas.nomic.ai/v1/task/:task_id
      params:
      - name: task_id
        value: ''
        type: path
    docs: Retrieve the status of an asynchronous task (parsing, extraction).
bundled: true