Trading increasingly spans multiple markets at once. PEXim gives you continuous, zero-risk evaluation with a zero look-ahead guarantee - settlement on real clearing prices, merit orders, and activations, not synthetic fills.
Connect your trading agent over REST or gRPC, place bids across markets, and receive benchmarked KPIs. One agent-id tracks your strategy end-to-end - from order entry through execution to P&L and leaderboard ranking.
Benchmarking mode. Asset owners invite multiple agents onto the same hybrid asset. Physical limits (MW, MWh, grid connection) are shared across agents and settlement; each agent runs its own schedule and keeps an independent SOC history.
Agents connect via REST or gRPC to the matching engine. The asset owner (right) supplies physical constraints and generation data to P&L settlement. Results flow back as executions, SOC, imbalance costs, and benchmarked KPIs.
Platform
A multimarket exchange in software
Power trading increasingly spans multiple markets at the same time. PEXim simulates day-ahead, intraday, and balancing products in one engine - with zero look-ahead and settlement on published market outcomes.
Unified API
One interface for day-ahead orders, balancing capacity and energy, and intraday continuous trading - plus queries for contracts, fills, and portfolio state.
Contract ledger
Every accepted bid becomes a contract with delivery windows, direction, and quantity. Positions roll forward quarter-hour by quarter-hour across markets.
Real-data settlement
Clearing and activation use published market outcomes - SDAC prices, merit orders, capacity awards, and intraday benchmarks - not synthetic random fills.
Markets
How each market is simulated
Products clear in sequence through the trading day. Later markets must respect commitments made earlier - capacity, day-ahead schedules, then intraday adjustments.
DA
Day-ahead auction
Agents submit limit orders via the API before gate closure. After the real SDAC clearing price is published, PEXim simulates execution: buy orders fill when the clearing price is at or below the limit, sell orders when it is at or above. Fills are at the published clearing price.
FCR
FCR capacity (planned Q4 2026, selected areas)
Frequency containment bids reserve MW at the edge of the flexibility envelope. Winning capacity obligations constrain what can still be traded in energy markets for each EFA block.
Cap
aFRR & mFRR capacity
Capacity bids are accepted through the API. After awards are known, agents must also submit the corresponding aFRR energy (and mFRR energy where applicable) placements. Acceptance follows published prices and quantities from the real capacity markets.
aFRR
Balancing energy & activations
Energy bids enter the merit stack. When real activation volumes and marginal prices are published, PEXim allocates activations across agents and settles at the published execution data - mirroring TSO dispatch logic.
ID
Intraday continuous
Intraday simulates against published benchmarks, tiered by subscription:
Starter - execution against quarter-hour VWAP curves.
Expert - real-time production replication for higher-fidelity fill simulation.
Adjustments stack on the day-ahead schedule and must stay inside the remaining flexibility envelope.
Assets
Physical limits and operator configuration
Asset owners register batteries, renewables, and flexible load with boundary conditions - power limits, energy capacity, grid connection constraints, and market eligibility per area.
Grid & nameplate limits
Each asset carries MW limits and optional ramp constraints. The simulator rejects or clips bids that would exceed configured capacity or violate grid export/import caps.
State of charge
Batteries track SOC across the delivery day. Schedules that would over-charge or over-discharge relative to SOC bounds are flagged and can trigger imbalance charges.
Integration
How your optimizer connects
Trading agents connect to PEXim via REST API or gRPC. One agent-id identifies your strategy across all markets, evaluation, and reporting.
Your Optimizer
agent-id: my-agent-01
Managed Assets
PV, BESS, Wind, Flex Load
REST / gRPC
Events / Streams
Generation actuals
PEXim Engine
DA, ID, aFRR, mFRR | FCR (Q4 2026)
Evaluation Backend
KPIs & Performance
Orders & book state
Submit bids and orders for all markets through a unified API. Query the order book, open positions, and fill history at any time via REST or gRPC.
Execution streams
Subscribe over gRPC to receive real-time messages on clearing results, capacity awards, activations, and intraday fills as they happen across all markets.
Agent evaluation
Each agent-id gets its own evaluation dashboard with P&L, VWAP deviation, imbalance costs, and performance timeseries - compare strategies side by side.
Generation
Asset output and generation feeds
PEXim tracks actual generation output alongside traded positions to measure real optimizer performance.
Actual generation feeds
Upload or stream actual generation output for managed assets - solar irradiance, wind capacity factors, or battery dispatch. PEXim compares net traded energy against what the asset actually delivered.
Portfolio-level view
For combined assets like PV+BESS, the evaluation backend tracks the aggregate portfolio: generation minus storage losses, net grid injection, and the resulting imbalance exposure per quarter-hour.
Settlement
Imbalance costs when physics diverges from trades
PEXim penalises agents that violate asset limits or deliver a different net profile than their combined market positions. Settlement follows real BRP/BSP perimeter logic: the agent bears imbalance responsibility for its traded schedule.
Power limit violation
A 10 MW battery whose net sold position totals 20 MW is short 10 MW of feasible delivery for as long as that mismatch persists. The engine records the imbalance volume and applies imbalance costs for the affected quarter-hours.
SOC violation
Similarly, dispatch that would require more energy than the battery can store or deliver triggers SOC violations and associated imbalance settlement.
Renewable generation profile
Solar and wind agents trade expected generation profiles. After each 15-minute period, when real output becomes available, any deviation between net traded quantity and actual feed-in creates imbalance exposure and cost.
Forecasts
Generation forecasts and fallback curves
Owners can inject asset-level generation forecasts and later upload realised production. PEXim uses these to judge whether traded schedules were achievable.
Owner-provided data
When forecasts and actuals are supplied, imbalance is measured against your asset’s own profile - the fairest basis for evaluating optimizer performance on renewables.
National fallback
Without asset-specific actuals, PEXim falls back to national average curves (e.g. typical PV or wind shapes for the area) and compares net traded energy to that reference, charging imbalance on the deviation.
Simulation
Why simulation matters
Testing on production markets is expensive, slow, and risky. Simulation solves all three.
1
Zero capital risk
Paper-trade on reality-calibrated market data without putting real money at risk. Backtest against historical data or run forward in real time. Deploy only when your algorithm proves itself.
2
Controlled Conditions
Compare algorithms on the same market data and matching rules. Fair benchmarking eliminates luck and market timing from the equation.
3
Faster Iteration
Run months of historical data in minutes. Replay specific market conditions to stress-test edge cases that rarely occur in live trading.
Onboarding
Getting started
Three paths into PEXim, depending on your role.
Optimizers
Create a trading agent, connect via the REST API, and start placing orders. Review fills, P&L, and VWAP deviation on your dashboard. Prove your route-to-market performance.
We support researchers, universities, and non-profit organizations who work to improve optimization models and publish their results openly for the benefit of the energy community. Special academic rates, shared knowledge, and consulting for those who contribute to open tools and public research.