Time-series foundation models as a service
One API. 49+ pretrained forecasters. No fine-tuning required.
TSFM.ai hosts every major open-weights time-series foundation model behind one consistent inference API — so you can benchmark, compare, and ship forecasts without managing GPU infrastructure or model weights.
curl -X POST https://api.tsfm.ai/v1/forecast \
-H "Authorization: Bearer $TSFM_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "amazon/chronos-2",
"inputs": [{"target": [10, 12, 11, 13, 14, 15, 14, 16, 18, 17]}],
"parameters": {"prediction_length": 24, "quantiles": [0.1, 0.5, 0.9]}
}'from tsfm import Tsfm
client = Tsfm()
forecast = client.forecast(
model="amazon/chronos-2",
inputs=[{"target": [10, 12, 11, 13, 14, 15, 14, 16, 18, 17]}],
parameters={"prediction_length": 24, "quantiles": [0.1, 0.5, 0.9]},
)
print(forecast.predictions[0].mean)49+ pretrained models from leading research labs, including:
| Family | Models |
|---|---|
| Chronos | Chronos, Chronos-Bolt, Chronos-2 |
| TimesFM | TimesFM 2.0, 2.5 |
| Moirai | Moirai 1.x, 2.0, MoE |
| Granite | Granite TTM |
| Others | TiRex, Toto, TimeMoE, MOMENT, Sundial, Timer, Lag-Llama, TEMPO, Kairos, YingLong, Kronos, TSPulse, PatchTST, FlowState, and more |
→ Full model list on Hugging Face
- GIFT-Eval — standardized evaluation across datasets and horizons
- Impermanent — live leaderboard tracking model performance over time
- 🌐 Website
- 📖 Documentation
- 📊 Blog
- 🤗 Hugging Face
- 💬 General: hello@tsfm.ai
- 🏢 Enterprise: sales@tsfm.ai