Shipped: 64,849 federal contract vehicles — GWACs, GSA Schedules, BPAs, queryable from your AI chat

If you do federal contracting work, you've heard the advice: "get on a vehicle." A Government-Wide Acquisition Contract (GWAC). A GSA Schedule. A Blanket Purchase Agreement (BPA). An Indefinite Delivery Contract (IDC). These are umbrella contracts that agencies issue task orders under — and getting on one is often a precondition for winning the work behind it.

But finding those vehicles, knowing who already holds them, when they expire, what NAICS codes they cover, and which set-asides they're eligible for — that has historically required either expensive subscriptions (GovWin IQ, GovTribe Scale tier) or weeks of FPDS / USAspending spelunking. Today we shipped contract-vehicle intelligence for BidSparq. Same flavor as our Beacon launch last week — ground-truth data, queryable from your AI chat, fully exposed via MCP, no bulk export. Here's the honest writeup.

What we shipped, in numbers

• 64,849 distinct federal contract vehicles (IDVs)
• 60,290 active (last_date_to_order in the future — orderable right now) = 93%
• 33,147 distinct holder companies (recipient_uei) — joins directly to our SAM-entity directory
• 52 awarding agencies — GSA leads at 21,435 vehicles (owns all FSS / GSA Schedules), DoD at 18,297, USDA at 6,487, VA at 5,579
• Full breakdown by type: 29,151 IDV_B_B (Single-Award IDCs), 15,488 IDV_C (BOAs), 9,296 IDV_E (Other), 5,742 IDV_B_C (FSS / GSA Schedules), 2,704 IDV_B_A (Multi-Award IDCs), 1,996 IDV_A (GWACs), 458 IDV_D (BPA parents)
• Sourced from USAspending.gov via their paginated /search/spending_by_award/ + per-award /awards/<id>/ detail endpoints — ground truth, not curated from a third-party
• NAICS, set-aside type (8(a) / HUBZone / WOSB / SDVOSB / SBA / etc.), pricing type (FFP / T&M / Cost-Plus), and multiple-vs-single-award flag populated on 99.995% of rows where USAspending has the data

How to use it

Three surfaces, same data:

1. AI chat — eight new tools

Eight new tools land in the BidSparq main chat (and the MCP server — accessible from Claude Desktop, Cursor, ChatGPT). The natural-language interface handles all of these:

• find_contract_vehicles — search by NAICS, agency, set-aside, vehicle type. "GWACs in NAICS 541512 that small businesses can compete on." "FSS schedules at GSA for IT services." "BPAs at the VA expiring in the next 12 months."
• get_vehicle_details — full profile of a single vehicle: holder, agency, period, NAICS, set-aside, vehicle type, pricing structure.
• get_vehicle_orders — task orders issued under a given vehicle.
• compare_vehicles — 2-5 vehicles side-by-side: ceiling, holder, expiration, ceiling utilization, sibling holders, competition tier. "Compare SEWP V vs Alliant 2 vs OASIS+." Pro Max tier.
• find_my_vehicle_opportunities — personalized. Uses the calling user's saved NAICS to recommend vehicles with enough ordering window remaining to be worth pursuing. Pro+ tier.
• vehicle_expiration_alerts — vehicles whose ordering window closes in the next N months. Strong recompete signal: agencies have to re-establish the vehicle to keep ordering. Pro+ tier.
• forecast_vehicle_recompete — for a single vehicle, projects the recompete window from its last-date-to-order plus typical federal lead times, with an action recommendation (prep now / monitor / too early). Pro+ tier.
• analyze_vehicle_competition — how crowded is a given vehicle. Multiple-vs-single-award structure, sibling holders, ceiling utilization, competition tier, bidder count. Pro Max tier.

2. Browse UI — /vehicles + /vehicles/[id]

The /vehicles index shows the top 25 active vehicles by award value. Every vehicle has a dedicated detail page at /vehicles/[id] with everything we know: holder, recipient UEI (joins to the vendor profile), agency hierarchy, NAICS, set-aside, vehicle type, ordering period, pricing structure, and linked task orders.

3. MCP server

All eight tools are exposed via our Streamable HTTP MCP server at https://bidsparq.com/mcp. From Claude Desktop, Cursor, ChatGPT, or any MCP-compatible AI client, your assistant can drive vehicle discovery without leaving its native UI. The npm package [email protected] is published, and the server is registered in the Official MCP Registry as com.bidsparq/[email protected].

What this is NOT

Being honest about scope so you can decide whether it solves your problem.

It's not a CPARS / past-performance database. We don't have ratings on how well the current holders performed. We know who holds the vehicle, not their evaluation scores. CPARS data is partially FOIA-gated and a separate ingest entirely.

It's not a GSA Advantage pricing tool. GSA Schedule prices for individual products are in GSA Advantage, which is a separate system requiring login. We know the vehicle exists, who holds it, and its NAICS coverage — not the SKU-level pricing.

It's not a recompete-prediction engine. We surface last_date_to_order as a strong recompete signal, but the actual prediction that "agency X will recompete vehicle Y in Q3 with these requirements" is inference territory we don't claim to do.

It's not state-level. USAspending is federal-only. State-level master contracts (like NASPO ValuePoint) come from different sources and aren't in this dataset.

And — same Beacon-style discipline — there is no bulk export. No CSV dump. No "list all vehicle holder emails" API. Vehicle data is exposed for research, not for outreach automation. Hard cap of 25 results per tool call. Privacy by design, not by oversight.

How this compares to GovTribe and GovWin

Direct comparison since the question always comes up.

GovTribe bundles federal IDV data and a contact graph into its Scale and Scale+ tiers, which require a sales call — pricing isn't public. BidSparq surfaces 64,849 federal vehicles at public Pro Max pricing ($149/mo, 14-day trial, no credit card required to start), queryable from your AI chat and any MCP client.

GovWin IQ sells federal pipeline intelligence including contract vehicles, typically at $15K–25K/year via annual contracts after a demo. BidSparq's vehicle data comes from the same public source (USAspending), with public pricing, full AI chat / MCP access, and 64,849 vehicles browseable without a sales call. If you're already paying for GovWin and using it heavily, keep it. If you're evaluating and want a lower-cost option that covers IDVs, this is one.

How to try it today

1. Visit bidsparq.com/vehicles — see the top 25 active vehicles by amount
2. From the BidSparq AI chat, ask things like "What GWACs are open in NAICS 541512?" or "Show me FSS schedules expiring in the next 12 months" or "Which vehicles should my company get on?"
3. From Claude Desktop, Cursor, or ChatGPT — if you have BidSparq's MCP server set up (install guide), the same eight tools are auto-exposed there
4. Open any vendor profile at /vendor/[uei] — if they hold contract vehicles, you'll see a new "Contract Vehicles Held" section

If you're already a Pro Max user, every surface is live now. Pro users get the search-tier tools (find_contract_vehicles, get_vehicle_details, get_vehicle_orders); the personalized and forecasting tools (find_my_vehicle_opportunities, vehicle_expiration_alerts, forecast_vehicle_recompete) are Pro+; and the competitive-analysis tools (analyze_vehicle_competition, compare_vehicles) are Pro Max.

If you're using BidSparq and federal vehicle data would change your workflow, we'd like to hear what you'd want next. Email [email protected] or just tell the chat.

Related reading

• BidSparq vs GovTribe — full side-by-side, including the vehicle / IDV coverage comparison
• Beacon — 85,978 procurement officers queryable from your AI chat — buyer-side counterpart to this release
• Using Claude or ChatGPT to monitor government RFPs — the MCP setup that exposes both vehicles and Beacon outside BidSparq
• How to spot a wired contract — combine vehicle holder intel with the 6 wired-risk signals
• All BidSparq features — vehicles in context with Beacon, Pursuits, MCP, and the rest of the platform