Plans
What's available on the Free and Pro tiers.
All accounts start on the Free tier. Free gives you access to pre-run baseline simulations so you can explore the data format and build integrations without needing to submit jobs. The Pro tier unlocks the full simulation engine — custom jobs, all scenarios, execution algorithm testing, liquidity-risk grids, validation, and foundation models.
Free
$0
- Access pre-run cached simulations
- Download simulation data as Polars DataFrames
- Bulk download multiple simulations as ZIP (3 new simulation groups / day)
- A selection of HKEX and LSE symbols, dates, and scenarios
- Up to 10 Monte Carlo runs per simulation group
Pro
Contact us
- Everything in Free
- Submit custom simulation jobs on any calibrated date
- All 6 market scenarios with custom parameters
- Execution algorithm testing (TWAP, VWAP, CSS)
- Liquidity-risk grids across a ladder of order sizes
- Pulse-Check validation of your runs
- Foundation-model inference on CPU with the public models; GPUs and further models enabled per account
- FIX 5.0 SP2 connectivity for your own execution algo
Free tier cached simulations cover a selection of HKEX and LSE symbols, calibration dates, and scenarios, with up to 10 Monte Carlo runs per group. See Simulations for how to access them.
Free-tier downloads are capped at 3 new simulation groups per rolling 24-hour period — but a group you have already downloaded stays free to re-fetch forever, in any file format. See Download limits for the details. Need more? Email support@simudyne.com and we can raise your download limit — no need to commit to the full Pro tier.
SDK methods by tier
| Method | Free | Pro | Description |
|---|---|---|---|
| simulation.list_cached() | Yes | Yes | List available pre-run simulations |
| simulation.get_sim_data() | Yes | Yes | Download one output file as a DataFrame |
| simulation.get_bulk_data() | Yes | Yes | Bulk download simulations as ZIP |
| simulation.get_sample_data() | Yes | Yes | Download a sample simulation without picking one |
| simulation.list_sim_files() | Yes | Yes | List the files available for one simulation |
| simulation.get_sim_params() | Yes | Yes | Get simulation configuration parameters |
| simulation.get_sim_metrics() | Yes | Yes | Get result metrics for one simulation |
| simulation.list_scenarios() | Yes | Yes | Scenario names and descriptions (local, no request) |
| simulation.get_scenario_defaults() | Yes | Yes | Default parameters for one scenario (local, no request) |
| data.calibrated_data() | Yes | Yes | List calibrated symbols and dates |
| data.available_data() | — | Yes | The data registry: every symbol-day that exists |
| data.calibrate() | — | Demo | Calibrate one symbol-day for the agent-based model |
| profile.get() / usage() / downloads() | Yes | Yes | Account, monthly request counts, download quota |
| api_keys.create() / list() / revoke() | Yes | Yes | API key management |
| fix.usage() | Yes | Yes | What you have run over a FIX session |
| simulation.run() | — | Yes | Submit a custom simulation job |
| simulation.run_lrm() | — | Yes | Market impact across a ladder of order sizes |
| simulation.get_jobs() | — | Yes | List your simulation jobs |
| simulation.get_job_status() | — | Yes | Track job progress, either engine |
| simulation.get_job_results() | — | Yes | Get results summary for a completed job |
| simulation.get_job_logs() | — | Yes | Fetch the run log for one of your jobs, either engine |
| validation.run() | — | Yes | Score simulated microstructure against the real session — platform runs, uploaded parquet files, or both |
| validation.get_job() / get_job_status() / list_jobs() | — | Yes | Fetch a result, poll a job, or list validation jobs |
| fm.models() | — | Yes | Discover foundation models and what each accepts |
| simulation.run(engine='fm') / fm.live() | — | Yes | Submit foundation-model inference, batch or streaming |
Every method above is available on the tier marked, with list_scenarios() and get_scenario_defaults() reading local constants rather than calling the API, so they work on any key. Validation is Pro-only and reports divergence only — L1 and Wasserstein distances across LOBbench microstructure metrics, a pass/fail verdict per Cont stylised fact, and the MIND and FID inception distances. All of these are aggregate scalars; the historical series itself is never returned. See Pulse-Check.
Endpoints without SDK methods
As of SDK 0.8.0 almost every documented endpoint has a method, so this list is short. Each of these is reachable with your X-API-Key header if you need them.
| Endpoint | Free | Pro | Description |
|---|---|---|---|
| GET /health | — | — | Liveness check. No authentication, no body |
| GET /fm/compute-profiles | — | Yes | The compute your account can run foundation models on (CPU plus any GPU types enabled for you), with approximate prices |
| POST /simulation/cached/{id}/download | Yes | Yes | Records a download from the web explorer against your quota — the SDK charges via the data endpoints instead, so you do not need to call this |
Every endpoint, its tier gate and its SDK equivalent are listed in the API reference.
Foundation-model compute
A Pro key runs the publicly available foundation models on CPU. Further models and GPU types (NVIDIA T4, L4 and A100) are enabled per account by Simudyne and billed per hour of compute. See Foundation models for the approximate prices and how to choose a GPU for a run.
Upgrading to Pro
To upgrade to Pro, email support@simudyne.com.