# shiyu-coder/Kronos

Kronos: A Foundation Model for the Language of Financial Markets

Repository: https://github.com/shiyu-coder/Kronos
Canonical: https://ross.abutalabs.com/products/kronos
Language: Python
License: MIT
License Family: permissive
Last push: 2026-04-13T12:38:49+00:00

## Health v2 (maintenance only)
Score: 53/100 (v2, computed 2026-09-02T17:46:02.011165+00:00)
- activity 77, release rhythm 35, longevity 30
- inputs: {"age_days": 428, "days_push": 142, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: no_releases
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 37962, forks 6312 (observed 2026-08-28T04:12:01.609801+00:00)

## What it is
Kronos is an open-source foundation model family for financial K-line (OHLCV) data, using a tokenizer plus autoregressive Transformer pre-trained on data from 45+ global exchanges. It provides a Python library with pretrained weights, a live demo, and fine-tuning scripts for quantitative finance tasks.

## Use cases
- predict stock price movements from candlestick data
- forecast crypto OHLCV time series
- fine-tune a foundation model on my own market data
- backtest quantitative trading signals
- model high-noise financial time series
- get K-line embeddings for downstream tasks

## When to choose
- you need a pretrained model specialized for financial candlestick data
- you want to fine-tune a market foundation model on your own exchange data
- you're doing quantitative research on OHLCV sequences

## When to avoid
- you need general-purpose time series forecasting outside finance
- you want guaranteed profitable trading advice - it's a research model
- you need a no-GPU lightweight solution

## Facets
- artifact type: library
- maturity: active
- function: machine-learning, deep-learning, llm-training, data-science, trading
- domain: fintech, machine-learning, deep-learning, large-language-models, time-series
- platform: python
- tags: foundation-model, financial-markets, k-line, ohlcv, time-series-forecasting, quantitative-finance, transformer, fine-tuning, gpu

## Member repositories
- shiyu-coder/Kronos (main) score 53

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:12:01.609801+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 2026-08-29T16:25:01.734395+00:00, confidence not recorded.
  - readme: https://github.com/shiyu-coder/Kronos (fetched 2026-08-28T04:12:01.609801+00:00, sha bb90a297981f)
- Data as of 2026-08-30T08:39:29.467469+00:00.