Popular Time Series Analysis AI tools

10 category leaders in Time Series Analysis, selected from the full directory.

Nixtla

  • Time-series AI platform behind TimeGPT foundation forecasting.
  • StatsForecast, MLForecast, and hierarchical forecasting toolkits.
  • 75M+ open-source downloads; $16M Series A in 2026.
  • Production zero-shot forecasting APIs plus open libraries.

GluonTS

  • Amazon-backed toolkit for probabilistic deep forecasting.
  • DeepAR, TCN, and Transformer-style time-series models.
  • Uncertainty quantification for univariate and multivariate series.
  • Default deep-learning stack for many forecast research pipelines.

Anodot

  • AI business monitoring and metric anomaly detection.
  • Learns normal time-series behavior across ops signals.
  • Glassbox company since November 2025 acquisition.
  • Enterprise TS anomaly detection for ops and finance.

TimeCopilot

  • GenAI agent that routes forecasts across 30+ TSFMs.
  • Natural-language forecasts, CV, and anomaly checks.
  • One API or CLI over Chronos, Moirai, TimesFM, TimeGPT.
  • Open agent layer for foundation-model forecasting.

TSFM

  • Unified API over major time series foundation models.
  • Zero-shot forecasts without hosting model weights.
  • Compare Chronos, TimesFM, Moirai, and more head-to-head.
  • Practical routing layer for production TSFM workloads.

Chronos

  • Amazon zero-shot time series foundation model.
  • Chronos-2 covers uni-, multi-, and covariate forecasts.
  • Hundreds of millions of Hugging Face downloads.
  • Default open TSFM for many AWS and research stacks.

TimesFM

  • Google Research foundation model for forecasting.
  • Native BigQuery AI.FORECAST without model training.
  • TimesFM-3 leads zero-shot GIFT-Eval and fev-bench.
  • Hyperscaler path for SQL-native time series AI.

Moirai

  • Salesforce universal zero-shot forecasting TSFM.
  • Handles any-variate series across domains and frequencies.
  • Moirai 2.0 improves GIFT-Eval accuracy and speed.
  • Open Uni2TS stack for research and production use.

Toto

  • Datadog multivariate TSFM for observability metrics.
  • Toto 2.0 family scales from 4M to 2.5B parameters.
  • Top-tier GIFT-Eval ranks; created the BOOM benchmark.
  • Open weights proving TSFM quality scales with size.

NXAI

  • European lab behind TiRex time-series world models.
  • xLSTM memory architectures for industrial forecasting.
  • TiRex-2 ranks top five on zero-shot GIFT-Eval.
  • Targets edge and embedded deployment beyond transformers.