$250 base + the AI you actually use.
Two pricing layers, nothing else. The base covers the operational BMS for unlimited users at $250/month flat. The AI layer bills on top, two modes — per-load ($1/load) or per-token consumption. Pick whichever maps to how your brokerage actually runs. Below: the cost math at four real brokerage shapes, side-by-side with the flat $997 TMS and the per-user incumbents.
| Shape | BMS · per-load | BMS · per-token | TMS flat | Aljex (per-user) | Recommend |
|---|---|---|---|---|---|
SMB starter 5 brokers · 50 loads / month | $250 + $50 = $300/mo | $250 + ~$40 = ~$290/mo | $997/mo | 5 × $499 = $2,495/mo | BMS per-load |
Mid SMB 10 brokers · 200 loads / month | $250 + $200 = $450/mo | $250 + ~$180 = ~$430/mo | $997/mo | 10 × $499 = $4,990/mo | BMS per-load |
Growing mid-market 20 brokers · 500 loads / month | $250 + $500 = $750/mo | $250 + ~$520 = ~$770/mo | $997/mo | 20 × $499 = $9,980/mo | Crossover — either BMS or TMS |
High-volume mid-market 30 brokers · 1,000 loads / month | $250 + $1,000 = $1,250/mo | $250 + ~$1,050 = ~$1,300/mo | $997/mo | 30 × $499 = $14,970/mo | TMS flat wins |
The three Keelway pricing tiers, plainly
The honest map of Keelway in 2026. Same product engine underneath all three — different ways to pay.
1 — Keelway AI · $1/load, no base
Pure inbox-AI layer that drops in on top of your existing TMS (Aljex, McLeod, AscendTMS, Tai). Best fit: shops happy with their TMS who just want the AI carrier-email triage. First 50 loads free.
2 — Keelway BMS · $250/mo base + flexible AI (this page)
The full broker management system at a predictable base, with AI metered. Best fit: shops looking to consolidate TMS + AI but with usage shape that's variable or genuinely unpredictable month-to-month. Below ~750 loads/month this is cheaper than the flat TMS; above that, the flat TMS wins.
3 — Keelway TMS · $997/mo flat, AI included
Full TMS with all AI usage bundled — no per-load, no per-token, no consumption tracking. Best fit: shops above ~750 loads/month with predictable volume who want one bill. Often the right destination from BMS after 2–3 months of usage data.
4 — Keelway Enterprise · custom
Dedicated tenant, SAML SSO, SCIM, signed MSA, 99.9% SLA, named CSM. For 200+ broker shops with procurement and security teams.
How to pick a tier in 90 seconds
- If you already have a TMS you like and just want the AI layer → Keelway AI ($1/load).
- If you're consolidating TMS + AI under 750 loads/month → Keelway BMS ($250 + flexible).
- If you're consolidating TMS + AI above 750 loads/month or want one predictable bill → Keelway TMS ($997 flat).
- If you're 200+ brokers with a procurement team → Keelway Enterprise.
Not sure? Start on BMS. The 30-day full-product trial lets you run real loads with real AI consumption and see which tier your actual usage maps to. Most brokerages know after week two.
Frequently asked questions
What's the exact breakeven between BMS and flat TMS?+
At $1/load AI, the BMS breakeven against the flat TMS is roughly 750 loads/month. Below 750 loads/month, BMS at $250 base + $1/load is cheaper than $997 flat. Above 750 loads/month, the flat TMS wins. Per-token consumption shifts the exact threshold up or down a bit depending on workflow (lighter AI workflows push the breakeven higher; heavier ones lower it), but $750 in monthly AI is the operative number to track. Most SMB brokerages under 500 loads/month stay on BMS; most above 1,000 move to TMS flat.
Are there any hidden charges?+
No setup fee. No per-user fee on the $250 base — that's the entire base. No annual contract requirement. The only line on the bill is base ($250) + AI consumption (your choice of mode), invoiced monthly. We'll send a usage projection in your dashboard so you can see the running consumption number any day of the month, not just at billing time.
Can I cap monthly AI spend?+
Yes. Set a hard monthly ceiling in the BMS billing settings — when consumption hits the cap, AI workflows pause for the remainder of the cycle (the operational TMS keeps running uninterrupted). Most brokerages don't use the cap; the ones that do typically set it 50% above their average month as a runaway-protection setting.
What if my AI usage is zero some months?+
Some brokerages have months — particularly during onboarding or off-season — where AI usage is near zero. In those months you pay the $250 base and that's it. There's no AI minimum.
What counts as one 'load' for per-load billing?+
A load is a unique load record in the BMS that runs through any AI workflow during its lifecycle. Posting a load and having the AI triage 30 carrier replies counts as one load. Running 5 check-calls on that load over its transit counts as the same one load. The $1 covers the load end-to-end. Loads that don't run through any AI workflow (e.g. a contract-lane dedicated load you handle manually) aren't charged.
How does per-token consumption pricing work?+
Token-based mode bills you for the actual LLM tokens consumed across the platform — both input and output, across all AI workflows on your account. Priced at our cost from Anthropic / OpenAI plus a small (~20%) margin. You see daily token consumption in the BMS billing dashboard. Most brokerages pick per-load for simplicity; per-token wins when your AI workflow shape varies a lot month to month.
Can I switch modes mid-cycle?+
Yes. The pricing mode is a setting in the BMS billing page — change it once per billing cycle. The change takes effect at the start of your next cycle. No contract lock-in, no penalty.