# Nixtla/nixtla

TimeGPT-1: production ready pre-trained Time Series Foundation Model  for forecasting and anomaly detection. Generative pretrained transformer for time series trained on over 100B data points. It's capable of accurately predicting various domains such as retail, electricity, finance, and IoT with just a few lines of code 🚀.

Repository: https://github.com/Nixtla/nixtla
Canonical: https://ross.abutalabs.com/products/nixtla
Homepage: https://www.nixtla.io/docs
Language: Jupyter Notebook
License: NOASSERTION
License Family: other
Topics: time-series, time-series-forecasting, deep-learning, forecasting, gpt, generative-ai-time-series, timegpt, anomaly-detection, artificial-intelligence, gpts, llm, foundation-models, llms, agentic-ai, agent
Last push: 2026-08-26T06:36:15+00:00

## Health v2 (maintenance only)
Score: 97/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 99, release rhythm 94, longevity 100
- inputs: {"age_days": 1805, "days_push": 7, "days_rel": 43, "gap_med": 29.0, "n_releases_24m": 13}
- flags: no_license
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 3996, forks 332 (observed 2026-08-28T04:08:31.887134+00:00)

## What it is
Nixtla's TimeGPT is a production-ready pre-trained foundation model for time series forecasting and anomaly detection, accessed via a Python SDK and hosted API (with Azure and self-hosted enterprise options). It performs zero-shot inference and fine-tuning across domains like retail, electricity, finance, and IoT.

## Use cases
- forecast future values of a time series with a few lines of code
- detect anomalies in historical or streaming time series data
- zero-shot forecasting without training a model per series
- fine-tune a pretrained forecasting model on my own dataset
- compare against ARIMA, LightGBM, or deep learning forecasters
- real-time online anomaly monitoring for metrics
- forecast retail sales, electricity demand, or financial data

## When to choose
- you need accurate forecasts quickly without per-series model training or tuning
- you want a hosted or self-hosted foundation model for time series with a simple Python API
- you need both forecasting and anomaly detection from one service
- you lack the compute or expertise to train deep learning forecasters

## When to avoid
- you need fully open-source weights or offline inference without an enterprise agreement
- you require strict data residency but cannot afford self-hosted deployment
- you need classical statistical forecasting with full interpretability
- your data is not time series (TimeGPT is not an LLM for text)

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, sdk, http-client, data-science
- domain: time-series, artificial-intelligence, data-science, analytics
- platform: python, cloud, self-hosted
- tags: timegpt, foundation-model, forecasting, anomaly-detection, zero-shot-inference, api-client, time-series-foundation-model

## Member repositories
- Nixtla/nixtla (main) score 97

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:08:31.887134+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-29T18:24:11.006183+00:00, confidence not recorded.
  - readme: https://github.com/Nixtla/nixtla (fetched 2026-08-28T04:08:31.887134+00:00, sha 153ed5c535dc)
  - homepage: https://www.nixtla.io/docs (fetched 2026-08-29T09:17:26.272253+00:00, sha 171c4f3a1bf3)
  - site_page: https://www.nixtla.io/docs/introduction/introduction (fetched 2026-08-29T09:17:26.276981+00:00, sha 171c4f3a1bf3)
  - site_page: https://www.nixtla.io/docs/introduction/why_timegpt (fetched 2026-08-29T09:17:26.278644+00:00, sha af9ac7479ad2)
  - site_page: https://www.nixtla.io/docs/introduction/about_timegpt (fetched 2026-08-29T09:17:26.280785+00:00, sha 8857f955f446)
  - site_page: https://www.nixtla.io/docs/introduction/timegpt_subscription_plans (fetched 2026-08-29T09:17:26.282722+00:00, sha 997ad84ff499)
  - site_page: https://www.nixtla.io/docs/introduction/faq (fetched 2026-08-29T09:17:26.284482+00:00, sha b511d25a5175)
  - site_page: https://www.nixtla.io/docs/setup/setting_up_your_api_key (fetched 2026-08-29T09:17:26.286561+00:00, sha c6773c313ea2)
  - site_page: https://www.nixtla.io/docs/setup/python_wheel (fetched 2026-08-29T09:17:26.288654+00:00, sha c54ddff6d359)
  - registry_pypi: https://pypi.org/pypi/nixtla/json (fetched 2026-08-29T09:17:26.290536+00:00, sha b82849ef23db)
- Data as of 2026-08-30T08:39:29.467469+00:00.