Lower offset from the reference centre
Institutional performance.
Built for better economics.
FX spreads of 0.2–0.3 pip in calm markets. Production pricing at 39.3 µs p99. Engineering-tested FIX TLS response at 20,000 orders/s. Real results behind better trading economics.
Commercial terms and leverage depend on the instrument and approved counterparty agreement. Slippage varies with market conditions, liquidity and order size.
PERFORMANCE & COST EFFICIENCY
Performance has a price.
Make it count.
Explore the cost of your trading flow, the capital it requires, and the engineering behind Drovix. Switch views to see the full picture.
39.3 µs. Measured in production.
Production · 05 Oct 2026Live pricing across FX, metals and crypto. The production engine delivered p99 tick-to-quote latency of 39.3 µs across all live symbols at approximately 569 decisions per second.
Singapore SG1 co-location, PTP-synchronised. Engine histogram, one-hour window ending before 00:09 UTC on 05 Oct 2026. The horizontal axis is categorical; lines connect measured checkpoints. This measures quote generation, not order throughput or external fill latency.
View all chart values & measurement notes
| Series | p50 | p99 | p99.9 |
|---|---|---|---|
| Drovix · all live symbols | 19.3 | 39.3 | 49.8 |
Drovix · all live symbols: Production engine histogram · FX, metals and crypto · approximately 569 decisions/s · Singapore SG1 co-location, PTP-synchronised.
MEASURED PRICING QUALITY / PRODUCTION
Tighter alignment.
Visible improvement.
Before → after the pricing change · 05 Oct 2026
Fewer quotes through an LP price
Drovix measurements: 05 Oct 2026 production and engineering results. Public terms reviewed 05 Oct 2026. Methodology & limitations ↗
Your trading economics.
Measured engineering results.
Cost-efficient liquidity. Capital flexibility. Execution designed to minimise slippage.
Drovix develops its own market-making and bridge technology. That gives technical teams a defined architecture to examine, from the pricing engine to the institutional interface.
Explore our technology0.2–0.3 pip FX spreads
Executable EURUSD and GBPUSD client quotes in calm markets. Gold at 15 points ($0.15). Observed in production on 05 Oct 2026; commission and financing depend on your agreement.
Discuss pricingHigh leverage
Discuss leverage and margin terms around your institution, instruments and risk profile. Available leverage is set in the approved agreement.
Discuss leverageSharper pricing
Gold centre offset p99 fell from $0.34 to $0.066 after the production pricing change. Quotes through an LP price fell from 20 to 2.95 per hour. These measure pricing quality; fill slippage is a separate metric.
Discuss executionPublished spreads and production metrics are achieved observations. Engineering tests report their own conditions below. Total trading cost also includes commission, financing and realised slippage.
Native market-making engine
Engine-to-bridge communication
Standards-based integration
The details
make the difference.
Eight broker-facing FX/CFD liquidity providers, selected for service overlap with Drovix. Public entry costs, margin terms and execution disclosures help frame your evaluation; company size is not a selection claim. Scroll the table to explore all eight providers.
Brand A–H represent real providers, kept anonymous. The source publications reviewed are listed below and available on request. Provider claims and measured Drovix measurements use different conditions; this table does not establish a performance ranking. “Not disclosed” means absent from the reviewed sources, not absent from the product.
| Technology & capability | DrovixMeasured results · 05 Oct 2026 | Brand A | Brand B | Brand C | Brand D | Brand E | Brand F | Brand G | Brand H |
|---|---|---|---|---|---|---|---|---|---|
| Commercial focusTrading costs | FX spreads 0.2–0.3 pipExecutable EURUSD/GBPUSD; gold $0.15 · calm markets · 05 Oct 2026 · spread only | Bridge includedNo extra MT4/MT5 bridge fee; all-in trading quote not published[A] | From $1,000/monthEntry offer; WL minimums differ. Spreads from 0.1 pip; not all-in[B] | FX spreads from zeroMinimum advertised spread; no comparable all-in tariff[C] | Tailored institutional termsIndicative spread feed; no comparable all-in tariff[D] | From $1,000/monthPublished entry fee; commissions and other costs separate[E] | Bespoke pricingNo comparable public all-in tariff[F] | From $7 / millionVolume tiers, monthly minimum and initial deposit apply[G] | Spreads from 0.0 pipsBespoke commissions; additional fees may apply[H] |
| Commercial focusLeverage & margin | High-leverage propositionRatio subject to instrument, risk approval and agreement | Confirm with providerNo numeric ratio in reviewed current offer[A] | Up to 1:200Subject to business needs and risk requirements[B] | FX margin from 1%Current global offer; instrument-dependent, not universal leverage[C] | 1:400 FXPublished headline; entity and instrument terms apply[D] | Tailored termsNo numerical institutional leverage in reviewed pages[E] | Cross-margin offeringNo comparable FX leverage ratio in reviewed pages[F] | Confirm with providerNo numeric ratio in reviewed launch disclosure[G] | Flexible marginNo numeric institutional ratio in reviewed page[H] |
| Commercial focusSlippage & fill quality | Low-slippage focusNo matched fill-slippage dataset presented; not established by ack latency | High fill-ratio claimProvider report on a single event; not a slippage distribution[A] | No last lookExecution policy claim; no matched slippage dataset[B] | Zero-slippage claimProvider marketing; no matched distribution supplied[C] | Execution consistency focusNo numeric slippage distribution in reviewed sources[D] | Flow analyticsNo matched slippage distribution published[E] | Low-slippage focusQualitative claim; no matched distribution[F] | Aggregated executionNo matched slippage distribution in reviewed disclosure[G] | Broad flow acceptanceNo matched slippage distribution published[H] |
| Service modelLiquidity relationship | Principal OTCBilateral institutional relationship | Prime of primeBroker FX/CFD liquidity[A] | Prime of primeProfessional B2B FX/CFD liquidity[B] | Aggregated CFD liquidityBroker-facing multi-asset offer[C] | Institutional FX/CFDsBroker, bank and asset-manager service[D] | Prime-of-prime servicesBrokerage firms, funds and trading desks[E] | OTC market making + DMABroker-facing multi-asset solutions[F] | Institutional aggregationBrokers, hedge funds and prop firms[G] | Institutional multi-assetBroker and professional-counterparty offering[H] |
| TechnologyEngineering approach | Owned engine + bridgeC++20 pricing core · Aeron IPC | Partner-built platformProprietary stack built by a technology partner[A] | Partner-built platformMatching, aggregation and risk controls[B] | Proprietary aggregation10-level market depth; bridge-free offer[C] | Integrated technology stackThird-party bridge and aggregation stack[D] | In-house technologyPosition-keeping and up to 10 levels of depth[E] | Proprietary aggregationMulti-platform technology ecosystem[F] | Third-party connectivity + co-locationNative integration and dedicated cross-connects[G] | ECN trading technologyLondon co-location connectivity[H] |
| ConnectivityAPI access | FIX 4.4Price and order sessions | FIX connectivityExecution, depth and hedge-account data[A] | FIX APIBroker platform connectivity[B] | Confirm API specificationReviewed offer does not specify protocol/version[C] | FIX APIDirect and bridge connectivity disclosed[D] | Confirm API specificationReviewed connectivity page focuses on MT4/MT5[E] | FIX APIThird-party trading access[F] | FIX APIListed institutional integration route[G] | FIX / REST / WebSocketPublished trading interfaces[H] |
| Different measurement boundariesReported latency¹ | 0.354 ms p99FIX TLS bridge response · 20,000/s · loopback engineering test · 05 Oct 2026 | Event average onlyProvider-reported for a single volatility event; no percentile[A] | Under 3 msExecution claim; percentile and test boundary unspecified[B] | Under 1 msExecution claim; percentile and test boundary unspecified[C] | Low latencyNo numeric value in reviewed current pages[D] | Low latencyNo numeric execution value in reviewed pages[E] | Low latencyNo numeric value in reviewed pages[F] | Hundredths of a msNetwork latency in some scenarios; not end-to-end fills[G] | Low latencyNo numeric execution value in reviewed page[H] |
| Evidence standardMatched cross-provider test | Not run for this page | Not run for this page | Not run for this page | Not run for this page | Not run for this page | Not run for this page | Not run for this page | Not run for this page | Not run for this page |
Real measurements.
Clearly defined paths.
Achieved results, reported 05 Oct 2026. Production observations and engineering tests are labelled at the measurement boundary.
All live symbols: FX, metals and crypto. p50 19.3 µs; p99.9 49.8 µs. Approximately 569 pricing decisions/s.
Singapore SG1 co-location · PTP-synchronised · engine histogram · one-hour window before 00:09 UTCAt 20,000 orders/s. At 30,000 orders/s sustained, p99 was 5.0 ms. Zero lost or duplicate orders.
Bridge inbound → 35=8 · Cloud Build loopback · tmpfs stores · 32 sessions · three alternated repetitionsZero refusals at 20,000, 30,000 and 40,000 orders/s and a 50,000 orders/s burst, with mixed order flow.
House-lane harness · real RocksDB WAL · 85 µs fsync emulation · engine-only boundary30,000 fills/s for 10 minutes, zero lost. Also 50,000 fills/s for 60 seconds. Booking lag p99 approximately 0.2 seconds.
Cloud Build · ClickHouse 26.4 · booking component, not end-to-end executionEURUSD and GBPUSD client quotes in calm markets. XAUUSD: 15 points ($0.15). Tight spreads delivered to clients.
05 Oct 2026 · calm-market observation · spread excludes commission and financingDown from $0.34 after the fair-value pricing change. Quotes through an LP price fell from 20/h to 2.95/h.
05 Oct 2026 · change from 00:32 UTC · pricing quality, not fill slippageThese are measured outcomes, not targets. Engineering throughput results have not been measured at these rates on production hardware. Pricing, bridge response, engine decisions and fill booking are distinct paths; their results do not combine into an end-to-end throughput claim or a service-level guarantee.
One connected
technology story.
Follow the responsibilities across the stack. Each layer has a distinct role in the institutional relationship.
Connect
Institutional interfacesFIX 4.4 price and order sessions connect the desk’s infrastructure to the agreed service scope.
Price & manage
Drovix market-makingThe C++20 engine brings pricing, inventory and risk decisions together, with Aeron communication between engine and bridge.
Review
Execution & reportingReview order states, execution records and reporting against the counterparty configuration and agreed measurement boundary.
Conceptual architecture. Available interfaces, configuration and service scope are established during institutional onboarding. Technology choices alone do not establish latency, fill quality or superiority over another provider.
A benchmark is only
as good as its method.
01Define the measurement boundary
Measure internal engine processing separately from client-to-gateway round-trip and external execution. Identify the start and end events, clock source and timestamp precision for every result.
02Match the test conditions
Use the same region, hardware class, network path, instrument set, message sizes and concurrency. Record the software version, warm-up period, test duration and offered load.
03Report the distribution
Publish p50, p95 and p99 latency alongside throughput, error rates and sample count. Include rejected instructions and timeouts so a fast subset cannot hide the complete outcome.
04Make the comparison reproducible
Record each provider’s configuration and data source. Preserve the test procedure and dated results. Anonymised Brand A–H results require the same evidence standard as Drovix.
¹ Latency figures in the comparison are not directly comparable. Brand A reports an average execution time for a single market event; Brands B and C advertise execution latency without a matched test boundary or percentile. Drovix reports local FIX TLS bridge response p99 in a loopback engineering test, not external trade fills. These values cannot establish a speed ranking. Entry fees and minimum spreads also do not establish all-in trading costs. Architecture references: Drovix technology, market-making and risk & controls. C++20 and Aeron are documented in the Drovix engine’s engineering documentation. Drovix results: production metrics and engineering test report dated 05 Oct 2026, supplied and approved for publication by Drovix. Exact conditions accompany each result. This page does not present an independent cross-provider performance study.
Follow the evidence.
Official provider publications support each comparison column. Providers are anonymised; the reviewed publications are available to counterparties on request. Published claims have not been independently re-tested.
Brand A
Liquidity offerTechnology & included bridgeDated execution and fill reportBrand B
FX/CFD pricing, leverage & execution offerService model & technologyBrand C
Current institutional liquidity, margin & execution offerBrand D
Institutional offer & FX leverageConnectivity & technologyBrand E
Institutional offer & monthly entry feeConnectivity & analyticsBrand F
Liquidity & broker servicesPlatforms & FIX APIBrand G
Institutional launch, fees & MT5 integrationBrand H
UK institutional offer & connectivityPut the right
questions to our desk.
Discuss a technical evaluation For approved institutional and professional counterparties.