{"id":678,"date":"2026-06-29T05:00:31","date_gmt":"2026-06-29T05:00:31","guid":{"rendered":"https:\/\/jua.ai\/articles\/europe-renewable-energy-forecast-api\/"},"modified":"2026-06-29T05:00:31","modified_gmt":"2026-06-29T05:00:31","slug":"europe-renewable-energy-forecast-api","status":"publish","type":"post","link":"https:\/\/jua.ai\/articles\/europe-renewable-energy-forecast-api\/","title":{"rendered":"Europe Renewable Energy Forecast API for Energy Desks"},"content":{"rendered":"<p><em>Written by: Olivier Lam, Physical AI Team, Jua.ai AG<\/em><\/p>\n<h2 id=\"key-takeaways\">Key Takeaways for European Energy Desks<\/h2>\n<ul>\n<li>European energy desks lose millions each year to forecast error because generic weather APIs and ENTSO-E data lack native renewable power outputs at trading cadence.<\/li>\n<li>Integrated physics-foundation-model APIs fix this gap by pairing high-accuracy atmospheric prediction with native power-output modeling, ensemble outputs, and fast refresh aligned with intraday markets.<\/li>\n<li>Jua for Energy delivers EPT-2 forecasts that outperform ECMWF HRES on every lead time, native solar, wind, and load forecasts for five European countries, and frequent actual-generation updates.<\/li>\n<li>Traders gain production-ready Python SDK access, live 25-model benchmarking, and Athena agent support that turns natural-language queries into briefings or backtests in under two minutes.<\/li>\n<li>Energy teams evaluating a Europe renewable energy forecast API should <a href=\"https:\/\/meetings-eu1.hubspot.com\/guett\/energy-trading?uuid=d780665f-ff71-439c-addf-c80e49af0627\" target=\"_blank\">book a demo with Jua<\/a> and run live benchmarks on their own region and variables in under five minutes.<\/li>\n<\/ul>\n<h2>The Problem: Generic Weather and ENTSO-E Data Miss Trading Needs<\/h2>\n<p>Generic weather APIs deliver atmospheric variables such as wind speed, irradiance, and temperature, but they do not provide renewable power outputs. Converting a wind-speed forecast at 100 m hub height into a generation forecast for a specific portfolio requires capacity data, turbine power curves, curtailment logic, and grid topology. Most desks either build that conversion layer themselves or skip it and trade on raw meteorological variables. Both approaches introduce avoidable error.<\/p>\n<p>The <a href=\"https:\/\/www.entsoe.eu\/data\/power-stats\/\" target=\"_blank\" rel=\"noindex nofollow\">ENTSO-E Transparency Platform<\/a> publishes actual generation and day-ahead forecasts for European balancing zones. Its data reflects what transmission system operators have already reported, not a forward-looking probabilistic view of generation in the next six hours. For intraday positioning, ENTSO-E behaves as a lagging indicator rather than a trading signal.<\/p>\n<p>A <a href=\"https:\/\/sciencedirect.com\/science\/article\/pii\/S0960148126008736\" target=\"_blank\" rel=\"noindex nofollow\"><em>Renewable Energy<\/em> study<\/a> quantifies the spatial structure of renewable forecast errors across Europe. Research shows that forecast errors for wind energy from models such as ECMWF IFS tend to become uncorrelated as site separation increases, and that spatial smoothing can reduce mean absolute errors and extreme forecast error events for wind and solar at day-ahead lead times. Because errors at different sites do not move in lockstep, a portfolio\u2019s total forecast error is not just the sum of individual site errors. Any API that treats each site independently will systematically underestimate the portfolio-level risk that traders actually face.<\/p>\n<p>The integrated physics-foundation-model API category addresses these gaps. It combines a physics-constrained atmospheric foundation model with native power-output modeling, ensemble probabilistic outputs, and a refresh cadence that matches the intraday market clock, not the HPC scheduling window of a national meteorological service.<\/p>\n<p style=\"text-align:center\"><strong>Validate this on your own book. Compare your region and variables head-to-head against 25+ models in under 5 minutes. <a href=\"https:\/\/athena.jua.ai\" target=\"_blank\">Run a live benchmark \u2192<\/a><\/strong><\/p>\n<h2>How Integrated Physics-Foundation-Model APIs Work in Practice<\/h2>\n<p>Jua is a foundation model and agent company, and Jua for Energy is its first applied product. The underlying technology is EPT (Earth Physics Transformer), a general spatiotemporal transformer foundation model that learns the governing physics of complex systems such as mass, momentum, and energy conservation directly from observational data. Athena is Jua\u2019s AI agent, instrumented with the Jua for Energy tool surface. The relationship mirrors Anthropic and Claude Code: a horizontal AI platform with a flagship vertical product.<\/p>\n<p>EPT-2, the flagship deterministic model, <a href=\"https:\/\/arxiv.org\/abs\/2507.09703\" target=\"_blank\" rel=\"noindex nofollow\">outperforms ECMWF HRES on every lead time across the full 0\u2013240 hour range<\/a> on 10 m wind, 100 m wind, 2 m temperature, and surface solar radiation. EPT-2e, the ensemble variant, <a href=\"https:\/\/arxiv.org\/abs\/2507.09703\" target=\"_blank\" rel=\"noindex nofollow\">beats the 50-member ECMWF ENS mean on both RMSE and CRPS at virtually every lead time<\/a>. Both results are benchmarked against more than 10,000 real ground stations on open-source StationBench, with no post-processing or station fine-tuning.<\/p>\n<p>A quant developer integrating Jua for Energy runs <code>pip install jua<\/code>, authenticates against the REST API, and pulls solar and wind generation forecasts for Germany, Great Britain, France, the Netherlands, or Belgium through a single schema. A typical workflow uses the EPT-2 wind-generation forecast for northern Germany at 06:00 UTC, compares it against the ECMWF HRES run on the same schema, flags the delta, and feeds both into the desk\u2019s intraday positioning model before the market opens. Athena runs that comparison and returns a written briefing in approximately 90 seconds.<\/p>\n<h2>Pain Point 1: Stale Forecasts Hurt Intraday Trading<\/h2>\n<p>Traditional numerical weather prediction (NWP) runs on HPC infrastructure that consumes approximately 8,400 kWh and costs \u20ac1,000\u2013\u20ac20,000 per simulation. The economics of that infrastructure cap update frequency at two to four global runs per day. Between runs, traders rely on stale numbers and react to weather only after it has already appeared in the price.<\/p>\n<p>EPT-2 RR, Jua\u2019s rapid-refresh model, updates up to 24 times per day. EPT-2 HRRR delivers the same hourly cadence at up to 5 km resolution over Europe. A single EPT-2 inference runs on a single GPU in minutes at approximately <a href=\"https:\/\/nebius.com\/customer-stories\/jua\" target=\"_blank\">0.25 kWh and $0.20\u2013$15 per simulation<\/a>, which is roughly four orders of magnitude cheaper than the equivalent NWP run. Actual-generation power forecasts on Jua for Energy refresh frequently throughout the day. The stale-data problem starts as a compute-economics problem, and the integrated physics-foundation-model API category solves it by replacing HPC with GPU inference.<\/p>\n<h2>Pain Point 2: Missing Native Power Outputs Create Risk<\/h2>\n<p>Generic weather APIs stop at atmospheric variables. Converting irradiance to solar generation, or 100 m wind speed to turbine output, requires a power-curve model, installed-capacity data, and curtailment logic that most weather API vendors do not provide. Desks that build this conversion layer internally take on a long-term maintenance burden. Desks that skip it trade on the wrong variable.<\/p>\n<p>Jua for Energy\u2019s Power Forecast surface covers solar, wind onshore, wind offshore, total wind, total renewables, load, and residual load in Germany, Great Britain, France, the Netherlands, and Belgium. Two complementary models run on the same surface. The Fundamental Model combines the EPT weather forecast with installed-capacity data and runs out to 20 days. The Actual Generation Model focuses on short horizons and updates throughout the day. Fundamental forecasts are capacity-weighted via Market Aggregates 2.0 with full cross-model comparison, deltas, disagreements, and heatmaps. Power output is native to the platform rather than translated from a generic weather variable after the fact.<\/p>\n<h2>Pain Point 3: Opaque Benchmarks Block Confident Procurement<\/h2>\n<p>Most energy desks evaluate forecast providers through vendor-supplied graphics or periodic accuracy reports. Neither approach is reproducible, and neither allows the desk to test the provider on the region and variable that actually drives their P&amp;L. Procurement decisions then rest on marketing materials instead of numbers.<\/p>\n<p>Jua for Energy\u2019s benchmarking surface puts more than 25 models on a single platform. It includes 10 proprietary AI models from the EPT family plus 15 third-party NWP and AI models such as ECMWF HRES, ECMWF ENS, ECMWF AIFS, NOAA GFS, GFS GraphCast, Microsoft Aurora, DWD ICON Global, and ICON-EU. A meteorologist evaluating Jua for Energy selects a high-stakes region and variable, runs the benchmark, and receives a head-to-head accuracy comparison in seconds. The benchmark is live, reproducible, and runs on the prospect\u2019s own inputs rather than on a vendor-curated test set.<\/p>\n<p style=\"text-align:center\"><strong>See the benchmarking surface on real data from your book. <a href=\"https:\/\/meetings-eu1.hubspot.com\/guett\/energy-trading?uuid=d780665f-ff71-439c-addf-c80e49af0627\" target=\"_blank\">Schedule a walkthrough \u2192<\/a><\/strong><\/p>\n<h2>Pain Point 4: High Integration Effort Slows Alpha<\/h2>\n<p>Quant teams that subscribe to AI weather research outputs such as DeepMind GraphCast, Microsoft Aurora, or ECMWF AIFS receive raw model files. Building the ingestion pipeline, ensemble logic, benchmarking harness, and hindcast access on top of those files consumes engineering capacity that should support alpha research. Integration that should take days stretches into a quarter.<\/p>\n<p>Jua for Energy exposes more than 25 models through a REST API (<code>POST \/v1\/forecast\/data<\/code> and related endpoints) with Apache Arrow support for large payloads. The Python SDK installs with a single command:<\/p>\n<pre><code>pip install jua<\/code><\/pre>\n<p>Hindcast data is available across multiple Jua and third-party models for backtesting. ENTSO-E grid-data integration supports European power-market data. The developer dashboard sits at <code>developer.jua.ai<\/code>, and full documentation at <code>docs.jua.ai<\/code>. Integration that takes a quant team a quarter to build elsewhere stands up in days.<\/p>\n<h2>Pain Point 5: From Raw Data to Tradeable View<\/h2>\n<p>Even desks with access to high-quality forecast data face a workflow bottleneck. Converting raw model outputs into a tradeable view of the day often requires manual assembly across terminals, spreadsheets, and vendor dashboards. Internal meteorology teams produce daily briefings by hand. External consultancies deliver reports after the trade window has closed.<\/p>\n<p>Athena, Jua\u2019s AI agent instrumented with the Jua for Energy tool surface, turns a natural-language objective into a briefing, a benchmark, a backtest, or a custom widget. A typical query such as \u201cwhat is the 100 m wind forecast spread across models for northern Germany tonight?\u201d resolves in under two minutes, as noted earlier. A backtest runs in approximately five minutes. Day-Ahead and Intraday briefings auto-refresh on every new model run and cover model consensus across more than 25 models, model delta since the previous run, convergence tracking, and price implications. <a href=\"https:\/\/nebius.com\/customer-stories\/jua\" target=\"_blank\">Trading houses and quant desks describe Athena as \u201canother headcount, for free.\u201d<\/a><\/p>\n<h2>Head-to-Head Comparison Table<\/h2>\n<table>\n<thead>\n<tr>\n<th>Attribute<\/th>\n<th>Jua for Energy (EPT family + Athena)<\/th>\n<th>Generic Weather APIs (e.g., Meteomatics, OpenWeather)<\/th>\n<th>ENTSO-E Transparency Platform<\/th>\n<th>AI Research Outputs (Aurora, GraphCast)<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td>Update cadence<\/td>\n<td><a href=\"https:\/\/arxiv.org\/abs\/2507.09703\" target=\"_blank\" rel=\"noindex nofollow\">Up to 24\u00d7\/day (EPT-2 RR); frequent actual-generation updates<\/a><\/td>\n<td>Hourly updates<\/td>\n<td>Varies by TSO; typically 15-min to hourly for actuals, day-ahead once daily<\/td>\n<td>Typically 4\u00d7\/day, no productised operational schedule<\/td>\n<\/tr>\n<tr>\n<td>Native renewable power forecast (DE, GB, FR, NL, BE)<\/td>\n<td><a href=\"https:\/\/nebius.com\/customer-stories\/jua\" target=\"_blank\">Solar, wind on\/offshore, load, residual load in 5 countries<\/a><\/td>\n<td>Atmospheric variables only, power conversion requires additional modeling<\/td>\n<td>Actual generation and TSO day-ahead forecasts, no probabilistic forward view<\/td>\n<td>Atmospheric variables only, no native power output<\/td>\n<\/tr>\n<tr>\n<td>Forecast horizon<\/td>\n<td><a href=\"https:\/\/arxiv.org\/abs\/2507.09703\" target=\"_blank\" rel=\"noindex nofollow\">Hourly to 20 days (deterministic), ensemble to 60 days<\/a><\/td>\n<td>Typically up to 7\u201316 days depending on tier<\/td>\n<td>Day-ahead only for forecasts, actuals in near-real-time<\/td>\n<td>Typically up to 10 days, research mode<\/td>\n<\/tr>\n<tr>\n<td>Python SDK and API quality<\/td>\n<td>REST plus Apache Arrow, <code>pip install jua<\/code>, hindcast access, single schema for 25+ models<\/td>\n<td>REST APIs available, schema varies by provider, hindcast access limited<\/td>\n<td>REST API available, no SDK, schema complexity high, no forecast hindcasts<\/td>\n<td>Research code or limited API, no productised SDK, no hindcast access<\/td>\n<\/tr>\n<tr>\n<td>Benchmark transparency<\/td>\n<td><a href=\"https:\/\/arxiv.org\/abs\/2507.09703\" target=\"_blank\" rel=\"noindex nofollow\">Peer-reviewed arXiv reports (2507.09703, 2410.15076), live 25-model benchmarking surface, StationBench against 10,000+ ground stations<\/a><\/td>\n<td>Vendor-supplied accuracy graphics, no independent peer-reviewed benchmarks<\/td>\n<td>Not applicable, actuals data not a forecast model<\/td>\n<td>Academic papers, no productised cross-vendor benchmarking surface<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<h2>Free and Paid Options for European Renewable Forecasts<\/h2>\n<p><strong>Free options.<\/strong> The <a href=\"https:\/\/transparency.entsoe.eu\/\" target=\"_blank\" rel=\"noindex nofollow\">ENTSO-E Transparency Platform<\/a> provides actual generation, installed capacity, and TSO day-ahead generation forecasts at no cost. It is the correct source for historical actuals and for understanding what TSOs have already reported. It does not provide a forward-looking probabilistic forecast and does not update at intraday trading cadence. NOAA GFS is available as a free deterministic baseline. It provides atmospheric variables but not native power outputs, and its accuracy on European wind and solar variables trails ECMWF HRES and EPT-2. Open-Meteo aggregates several free NWP outputs through a public API with query restrictions.<\/p>\n<p><strong>Paid integrated physics-foundation-model APIs.<\/strong> Jua for Energy is the production-ready option in this category. It combines EPT-2\u2019s superior atmospheric forecasts with native renewable power outputs, a 20-day fundamental forecast horizon, ensemble probabilistic outputs via EPT-2e, and a Python SDK with hindcast access. The Athena agent layer adds natural-language briefings, benchmarks, and backtests. Pricing is available on request, and a benchmark proof-of-value runs in under five minutes through the Athena interface.<\/p>\n<h2>Risks and Due Diligence for Forecast API Selection<\/h2>\n<p>Any evaluation of a Europe renewable energy forecast API should apply the following criteria before procurement.<\/p>\n<p><strong>Live benchmarking capability.<\/strong> Vendors that cannot provide a reproducible, live benchmark on the buyer\u2019s own region and variable should be treated with caution. Vendor-supplied accuracy graphics do not replace a head-to-head comparison on the buyer\u2019s highest-stakes input. The benchmark should run against a recognized reference such as ECMWF HRES or ECMWF ENS, and the methodology should be documented in a peer-reviewed source.<\/p>\n<p><strong>Schema stability.<\/strong> Quant teams and engineering desks building systematic strategies on top of forecast APIs cannot tolerate schema changes that break pipelines. Evaluate the API\u2019s versioning policy, the quality of its OpenAPI documentation, and whether it supports Apache Arrow for large-payload queries before committing to integration.<\/p>\n<p><strong>ENTSO-E integration.<\/strong> Native renewable power forecasts that cannot be reconciled against ENTSO-E actual generation data are difficult to validate in production. Confirm that the API provides a direct ENTSO-E integration or a documented methodology for aligning forecast outputs with TSO-reported actuals. A <a href=\"https:\/\/sciencedirect.com\/science\/article\/pii\/S0960148126008736\" target=\"_blank\" rel=\"noindex nofollow\"><em>Renewable Energy<\/em> study (2026)<\/a> confirms that spatial error dependence in European renewable forecasts is non-trivial. Any API that cannot model portfolio-level spatial aggregation will underestimate imbalance risk.<\/p>\n<p><strong>Hindcast availability.<\/strong> Backtesting a systematic strategy requires years of historical forecast data at the same schema as the live feed. Confirm hindcast availability, coverage period, and whether hindcasts are available for third-party models on the same platform before signing.<\/p>\n<p style=\"text-align:center\"><strong>Validate the claims with your own data. <a href=\"https:\/\/athena.jua.ai\" target=\"_blank\">Start your benchmark in under 5 minutes \u2192<\/a><\/strong><\/p>\n<h2>Frequently Asked Questions<\/h2>\n<h3>What defines an integrated physics-foundation-model renewable energy forecast API?<\/h3>\n<p>An integrated physics-foundation-model renewable energy forecast API combines three capabilities that generic weather APIs and raw NWP feeds do not provide together. It uses a physics-constrained atmospheric foundation model that produces high-accuracy meteorological variables. It includes a native power-output modeling layer that converts those variables into renewable generation forecasts without requiring the buyer to build the conversion logic. It also offers programmatic API access with ensemble probabilistic outputs, hindcast availability, and a refresh cadence that matches intraday market windows.<\/p>\n<p>The physics-constrained requirement matters because a standard transformer applied naively to atmospheric data can produce outputs that violate conservation laws. A physics foundation model like EPT learns mass, momentum, and energy conservation directly from observational data, so its outputs remain physically constrained by construction. The peer-reviewed validation requirement distinguishes this category from vendor-supplied accuracy claims. Benchmark results should be reproducible against recognized references such as ECMWF HRES, documented in arXiv technical reports, and testable by the buyer on their own region and variable.<\/p>\n<h3>How should performance of a Europe renewable energy forecast API be evaluated?<\/h3>\n<p>Performance evaluation should cover four dimensions. First, measure deterministic accuracy using RMSE and MAE against ground-truth observations, not model analyses, on the variables that drive the buyer\u2019s P&amp;L such as 100 m wind speed for wind generation, surface solar radiation for solar generation, and 2 m temperature for load. The benchmark should cover the full forecast horizon relevant to the buyer\u2019s trade windows, typically 0\u201348 hours for intraday and day-ahead and 3\u201310 days for multi-day positioning.<\/p>\n<p>Second, measure probabilistic skill using CRPS on ensemble outputs, evaluated against the ECMWF ENS mean as the reference. Third, review update cadence by counting how many times per day the forecast refreshes and checking dissemination latency relative to the underlying model run. Fourth, assess native power-output accuracy. The forecast should be evaluated on generation megawatts, not just on atmospheric variables, using ENTSO-E actual generation as the ground truth. A live benchmark that covers all four dimensions on the buyer\u2019s own region provides the most reliable evaluation method.<\/p>\n<h3>What are the typical integration steps for a renewable energy forecast API in Europe?<\/h3>\n<p>Integration typically follows four steps. First, install the SDK and authenticate. For Jua for Energy, this means running <code>pip install jua<\/code> and configuring the API key through the developer dashboard at <code>developer.jua.ai<\/code>. Second, become familiar with the schema. Review the OpenAPI documentation at <code>query.jua.ai\/docs<\/code> to understand the forecast endpoint structure, variable naming conventions, and payload format. Apache Arrow support is available for large multi-model, multi-variable queries.<\/p>\n<p>Third, pull hindcasts and run a backtest. Retrieve historical forecast data for the target region and variable, align it against ENTSO-E actual generation using the platform\u2019s direct ENTSO-E integration, and validate forecast accuracy against the buyer\u2019s own ground truth before going live. Fourth, integrate the live feed. Pipe the live forecast endpoint into the desk\u2019s existing trading, risk, or dispatch system. For desks that prefer a natural-language interface, Athena can generate custom widgets and dashboards on request in approximately 90 seconds without requiring engineering work.<\/p>\n<h3>Who typically uses a solar or wind forecast API in Europe?<\/h3>\n<p>Three buyer archetypes use renewable energy forecast APIs in European markets. Regulated utilities such as EDF, EnBW, Statkraft, and Axpo use solar and wind generation forecasts for dispatch decisions, balancing-responsible-party obligation management, and day-ahead trading. Their evaluation is led by internal meteorologists who benchmark forecast accuracy against ECMWF HRES on the utility\u2019s own generation portfolio.<\/p>\n<p>Physical trading houses, including commodity desks at companies like TotalEnergies, use renewable generation forecasts as a primary input to intraday and day-ahead power positioning. Their evaluation is fast and benchmark-driven. They run a live accuracy comparison on their highest-stakes region and variable and procure within weeks if the numbers are credible. Capital-markets and quantitative funds use renewable generation forecasts as one of several systematic signals fed into algorithmic strategies. Their evaluation is entirely programmatic. They install the SDK, run backtests against years of historical forecasts, and decide based on out-of-sample accuracy. All three archetypes are active users of Jua for Energy across five continents.<\/p>\n<h2>Conclusion: Why Jua for Energy Fits Trading Workflows<\/h2>\n<p>Generic weather APIs deliver atmospheric variables without native power outputs. ENTSO-E delivers actuals without a forward-looking probabilistic view. AI research outputs deliver raw model files without ensembles, hindcasts, or a productised refresh schedule. None of these options provides a Europe renewable energy forecast API built specifically for trading.<\/p>\n<p>The integrated physics-foundation-model API category closes these gaps. It offers physics-constrained atmospheric accuracy validated by peer-reviewed benchmarks, native renewable power outputs that update throughout the day, a 20-day fundamental forecast horizon, ensemble probabilistic outputs that beat the ECMWF ENS mean, and a Python SDK that stands up in days rather than a quarter. Jua for Energy is the production-ready implementation of that category, used by Axpo, TotalEnergies, Statkraft, EnBW, EDF, and quant funds across five continents.<\/p>\n<p>The benchmark provides the proof. Run it on your own region and variables in a head-to-head comparison against more than 25 models in under five minutes. <a href=\"https:\/\/athena.jua.ai\" target=\"_blank\">Run your benchmark now \u2192<\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Generic APIs and ENTSO-E fall short for energy trading. Jua delivers high-accuracy renewable forecasts across Europe. Start your free trial today.<\/p>\n","protected":false},"author":103,"featured_media":677,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"inline_featured_image":false,"footnotes":""},"categories":[13],"tags":[],"class_list":["post-678","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-product"],"_links":{"self":[{"href":"https:\/\/jua.ai\/articles\/wp-json\/wp\/v2\/posts\/678","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/jua.ai\/articles\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/jua.ai\/articles\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/jua.ai\/articles\/wp-json\/wp\/v2\/users\/103"}],"replies":[{"embeddable":true,"href":"https:\/\/jua.ai\/articles\/wp-json\/wp\/v2\/comments?post=678"}],"version-history":[{"count":0,"href":"https:\/\/jua.ai\/articles\/wp-json\/wp\/v2\/posts\/678\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/jua.ai\/articles\/wp-json\/wp\/v2\/media\/677"}],"wp:attachment":[{"href":"https:\/\/jua.ai\/articles\/wp-json\/wp\/v2\/media?parent=678"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/jua.ai\/articles\/wp-json\/wp\/v2\/categories?post=678"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/jua.ai\/articles\/wp-json\/wp\/v2\/tags?post=678"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}