The CQG alternative for futures brokers building an AI-native execution stack.
CQG is genuinely one of the best futures execution stacks in the world. It is also a per-user, plugin-constrained architecture designed for professional traders — not for a broker that wants to reason across every client, every fill and every risk event with modern AI.
1 — What CQG actually is
CQG has been building professional-grade futures technology since the 1980s. Its stack today spans four core products: CQG Integrated Client (the professional-trader workstation), CQG QTrader (a lighter execution client), CQG One (retail-friendly desktop), and CQG Desktop (browser-based). Underneath sits FIX connectivity to 45+ exchanges, a market-data infrastructure across 85+ sources, and a well-regarded API surface.
For serious futures traders — prop shops, commodity hedgers, CTAs, professional retail — CQG is a genuinely differentiated product. Depth of exchange coverage, quality of market data, execution reliability and the analytics footprint are all top-tier. Brokers who serve professional futures clients typically offer CQG because their clients ask for it by name.
The question — as with every serious platform vendor — is not whether CQG is good. It is whether the per-user pricing, the professional-trader-centric UX and the plugin/API model are the right architecture for a broker that wants to build firm-wide AI systems on top of everything CQG generates.
2 — What CQG costs you
CQG publishes its pricing openly, which is genuinely rare in this market. The core numbers for a broker to model, taken from CQG's own monthly-fees pages.
| Line item | Order-of-magnitude cost | Note |
|---|---|---|
| CQG Integrated Client (base) | $595 / user / month | Professional workstation; includes analytics + RTD. |
| CQG Trading add-on (Integrated Client) | +$250 / user / month | Execution rights on top of the base package. |
| CQG Spreader | +$1,400 / user / month | Advanced spread trading module. |
| CQG QTrader / CQG One | $100 base + $0.25 / fill (cap $1,000) | Volume-based; suits active retail / semi-pro. |
| CQG Desktop | $25 base + $0.25 / fill (cap $695) | Browser-based, retail-friendly. |
| API add-ons (Real-Time, Trading, Historical) | $45 – $245+ / month | Historical packages start at $100. |
| Exchange fees + pro/non-pro data fees | Passed through per exchange | Meaningful additional line at scale. |
For a broker with 400 professional Integrated Client users all on the trading package, the direct CQG line lands around $4.0m a year before API add-ons, exchange fees or the loaded cost of dealing-desk and back-office integration. That number is entirely defensible for the value CQG delivers to those users — the question is what happens above that spend.
3 — What CQG can't do for your firm
CQG is a professional futures execution stack. It is not, and does not try to be, a firm-wide AI operating layer. Five gaps that recur in the futures and multi-asset brokers we speak to.
It cannot reason across every client, every fill and every risk event
CQG sees orders, fills and market data. It does not see the CRM, the deposit history, the margin call thread, the support ticket, the sales-desk voice memo. The broker's most valuable AI system — predicting margin stress and dormancy across the entire book — needs all of that. CQG is one input, not the platform.
Per-user pricing punishes AI-augmented workflows
The whole point of an AI-native broker is that one dealing-desk operator can supervise the work of five. Under a per-seat vendor model, saving five FTEs of manual effort also saves five seats of vendor licence — a cost the vendor has every incentive to keep charging. Firm-owned intelligence layers make that a broker's saving, not a vendor's revenue.
The API surface is powerful but siloed
CQG's APIs are genuinely good, but they are CQG-shaped. Building a firm's cross-vendor AI on top means every model has to speak CQG on one side and something else on the other. A firm-owned data spine that all vendors feed into is a much better long-term shape.
Data ownership is one-way
CQG will happily stream data out via its APIs. It is much less interested in accepting AI outputs (scoring, risk overlays, retention triggers) back into the professional-trader UI. The intelligence layer therefore has to live outside CQG, which is fine — but it needs to be built.
Roadmap is set by CQG's professional-trader audience
CQG is a great vendor for its target user — the professional futures trader. Broker-side workflows (dealing desk, retention, compliance, margin surveillance) are secondary to that audience and will always be a smaller part of the roadmap. Brokers who want to move on those workflows have to own them.
4 — Why 'your own version' is viable in 2026
Five years ago, the idea of a firm-owned AI layer around a stack like CQG was optimistic. In 2026 it is the direction serious futures and multi-asset brokers are already moving in — quietly, because their edge depends on you not knowing.
The reason it works now is that the pieces are finally mature. FIX-out from CQG (and every other execution venue) into a governed warehouse is well understood. Modern reasoning models are genuinely good at margin-stress, dormancy and surveillance workloads when grounded on the right data. The engineering toolchain — streaming pipelines, vector infra, orchestration, evaluation — no longer needs a 30-person platform team.
You are not replacing CQG. You are building the layer that turns CQG's fills, market data and risk events into decisions faster than your competitors can make them. In that architecture, CQG remains the professional-trader execution stack. Your intelligence layer becomes the firm-owned brain around it.
5 — Reference architecture
Firm-owned intelligence layer around CQG and every other execution and back-office vendor
Professional traders keep CQG. The Integrated Client, the Spreader, the API-driven workflows — none of that changes at the terminal layer. What changes is what happens behind the broker.
Every fill, every position change, every risk event streams out of CQG (and out of every other execution venue the broker runs) into a firm-owned data spine that combines with CRM, deposits, support and marketing data. On that spine, a small AI team can build things CQG will never build: firm-wide margin-stress prediction, dormancy and reactivation agents, cross-venue surveillance, IB analytics.
The broker's dealing desk, risk and retention cockpits read from the same spine, so the whole firm sees one version of the client. CQG remains best-in-class where it is best in class. Everything above the fill becomes firm-owned.
6 — Build vs. rent: 3-year TCO
A three-year illustrative comparison for a mid-sized futures broker with 400 professional users on CQG Integrated Client + Trading.
| Dimension | Rent (CQG) | Build (KJ Capital) |
|---|---|---|
| CQG Integrated Client + Trading (400 users) | ≈ $4.05m / yr | ≈ $4.05m / yr (unchanged) |
| API + historical data add-ons | $150k – $400k / yr | $150k – $400k / yr |
| Third-party risk + surveillance vendors | $400k – $900k / yr | $150k – $300k / yr (fewer needed) |
| Intelligence layer (LLM infra + evals) | N/A | $300k – $500k / yr |
| Initial build (Yr 1 only) | N/A | £400k – £800k one-off KJ Capital build |
| 3-year total (illustrative) | ≈ $14m of spend, no owned asset | ≈ $14m of spend + a compounding proprietary layer |
The point is not to save money on CQG. The point is that the money the broker is already spending buys a professional-trader terminal — not a firm-wide AI operating layer. The intelligence layer is what closes that gap, at roughly the incremental cost of one or two additional risk or surveillance vendors.
7 — The three honest risks of building your own
"Our clients ask for CQG by name — we can't touch the terminal."
How we solve it —We don't. The intelligence layer is entirely broker-side. Professional traders keep the exact CQG configuration they know, with zero disruption. The value we add is behind the desk — for the dealing desk, risk, compliance and retention teams.
"CQG's API surface is stable — why not build inside it?"
How we solve it —For a single narrow tool inside CQG, that is fine. For a firm-wide intelligence layer that reads from CQG plus CRM plus payments plus a second execution venue, the layer has to live outside any single vendor. Otherwise every model is coupled to CQG's roadmap.
"We don't have a data or AI team big enough to run this."
How we solve it —You don't need one on day one. The Build engagement delivers the initial system in 6–12 weeks. Our Operator engagement runs it in production while your team takes ownership — the goal is that you are independent within 18 months.
8 — The CTA ladder
CQG will keep being one of the best futures execution stacks in the world. That is not the question. The question is whether the broker-side layer above CQG is going to be a spreadsheet, a BI dashboard, and a handful of manual workflows — or a firm-owned intelligence layer that compounds returns every quarter it runs.
The right sequence is Readiness → Audit → Blueprint. Two weeks with our team gives you a scoped plan you can put in front of your CEO and Head of Risk.
Free Readiness Score
5 minutes. Honest score of whether your firm is in a position to build.
AI Diagnostic — £15k / 2 weeks
Founder-led. Map the specific systems where you'd get compounding return.
Blueprint conversation
30-minute call with Kasim to scope the Build.
Five questions we get asked most.
Are you saying futures brokers should stop using CQG?+
No. CQG remains best-in-class for professional futures execution and its clients ask for it by name. What we advocate is building a firm-owned intelligence layer around CQG (and every other execution venue) that turns fills, risk events and client behaviour into broker-wide decisions modern AI can support.
How much does CQG actually cost per user?+
Per CQG's own published pricing: $595/month for the Integrated Client base package, plus $250/month for CQG Trading rights, plus $1,400/month for the Spreader module. Lighter clients (QTrader, CQG One, Desktop) are transaction-based, typically $100 base + $0.25/fill up to a $1,000 cap.
Can we build the intelligence layer inside CQG's APIs?+
For narrow, terminal-side tools yes. For a firm-wide layer that reasons across CQG plus CRM plus payments plus other execution venues, the layer must live outside any single vendor — on a firm-owned data spine that all vendors feed into.
What is the highest-ROI first system for a futures broker?+
In our experience, firm-wide margin-stress prediction plus a proactive risk cockpit is the highest-ROI first system for a serious futures broker. It compounds directly into reduced defaults, better client experience and lower manual dealing-desk load.
How long does an initial build take?+
The Diagnostic runs for two weeks. The initial Build engagement is 6–12 weeks and typically delivers one production system end to end — usually margin stress or dormancy. Operator runs it in production while the broker's team takes ownership.