Amazon logo

Chronos-Bolt (Mini)

online
amazon/chronos-bolt-mini

21M params | 2K context | $0.00025 per forecast

Chronos-Bolt Mini is a compact 21M-parameter checkpoint in Amazon's Chronos-Bolt family. It sits one rung above Chronos-Bolt Tiny, trading a little extra capacity for a better accuracy-latency balance on moderate-volume forecasting workloads while staying firmly in the cost-efficient tier of the line.

Like the rest of Chronos-Bolt, it runs a patch-based T5 encoder-decoder and produces forecasts through direct multi-step quantile prediction rather than autoregressive token-by-token rollout, so it returns probabilistic forecasts in a single pass without the latency of the original Chronos generation. It is pretrained on nearly 100B time-series observations drawn from large public corpora and synthetic pretraining data, as documented by the official Chronos-Bolt model cards.

On TSFM.ai, Mini is the pick when Tiny feels too lean but you still want fast, low-cost zero-shot quantile forecasts at scale. Step up to Chronos-Bolt Small or Base when accuracy matters more than per-call cost, or move to Chronos-2 when your workload needs covariates or multivariate cross-learning rather than lightweight univariate throughput.

Model Classification

Family

Chronos

Type

time series foundation model

Pretrained time-series model exposed on TSFM.ai for zero-shot or few-shot forecasting workloads.

Training Data

Nearly 100B time-series observations from large public corpora and synthetic pretraining data, as documented by the official Chronos-Bolt model cards.

Recommended For

  • Fast zero-shot probabilistic forecasting in production APIs
  • Teams replacing classic Chronos-T5 with lower latency inference

Strengths

  • Direct multi-step forecasting avoids autoregressive rollout cost
  • Strong default for quantile forecasts with lightweight serving

Limitations

  • Focused on forecasting rather than broader multi-task time-series work
  • Less natural fit when you need model behavior tuned around rich multivariate covariates

Capabilities

forecastingquantile-forecastingzero-shothigh-throughput

Tags

amazonchronosprobabilisticfastcost-efficient

Specifications

Parameters
21M
Architecture
patch-based T5 encoder-decoder with direct multi-step quantile forecasting
Context length
2,048
Max context
2,048
Minimum history
n/a
Recommended history
n/a
Input step
n/a
Required target series
1
Temperature
Ignored
Top P
Ignored
Max output
1,024
Avg latency
n/a
Uptime
n/a
Plan limits
1,000 rpm free · 1,000,000 rpm with billing
Accelerator
T4
Regions
Virginia, US
License
n/a

Pricing

Per forecast
$0.00025

Performance

Average latency
n/a
Availability
n/a
Plan limits
1,000 rpm free · 1,000,000 rpm with billing

Related models