See the whole monthly cost before you choose.
Choose one company tier: Essential is $50/month + $0.01 per BrokerBot-sent message, Pro is $150/month + $0.02, and Max is $400/month + $0.05. Every tier includes all loan officers; model processing is shown separately in the estimates.
How many messages will BrokerBot send?
The totals below include the company subscription, message charges, and estimated model processing. Applicable taxes are not included.
Three tiers, one complete platform. Processing for Essential and Pro is charged through BrokerBot. Max uses your company’s Anthropic key, so its processing is billed by Anthropic. The Max estimate combines both bills to keep the full expected spend visible.
Essential
Fastest OpenAI option in BrokerBot’s same-day evaluation, with a 4.2-second median reply and a 0.7% measured conversation-violation rate.
- Best fit for most companies
- Lowest measured OpenAI compliance risk
- No separate model-provider account
- Full product and all officers included
Pro
The most consistent OpenAI option in the internal evaluation, with only 8.8% of texts scoring under 4.
- Best when the worst reply matters most
- More consistent response quality
- No separate model-provider account
- Full product and all officers included
Max
Strongest opt-out handling in the internal evaluation, with a 5.9% measured conversation-violation rate.
- Best for compliance-first teams
- Your company owns the Anthropic billing relationship
- Requires a valid company Anthropic key
- Full product and all officers included
Calculator assumptions: all dollar amounts are USD. Every displayed message uses the model shown, at measured caching-aware processing averages of $0.005040 for Essential, $0.038703 for Pro, and $0.214245 for Max. Actual processing follows token usage, may include other model work, and can vary by conversation or model mix. Estimates exclude applicable taxes, Bonzo, and other third-party services. A signed order controls if it states different commercial terms.
What changes—and what doesn’t.
Every tier includes the complete BrokerBot platform and every loan officer. The monthly base, assistant-message rate, processing cost, provider, and model behavior change by tier.
| Capability | Essential | Pro | Max |
|---|---|---|---|
| AI model | GPT-5.6 Luna | GPT-5.5 | Claude Fable 5 |
| Company tier base / month | $50 | $150 | $400 |
| Additional officer platform fee | $0 | $0 | $0 |
| Assistant-sent message charge | $0.01 each | $0.02 each | $0.05 each |
| Measured processing estimate / assistant message | $0.005040 | $0.038703 | $0.214245 |
| Who bills model processing? | BrokerBot | BrokerBot | Anthropic · paid directly |
| Full Bonzo sync and history | Included | Included | Included |
| Pipeline and stage behavior controls | Included | Included | Included |
| Per-officer assistants | Included | Included | Included |
| Public + private testing ground | Included | Included | Included |
| Best internal benchmark signal | Speed and low violation rate | Consistency | Opt-out handling |
Useful evidence, clearly bounded.
The July 2026 BrokerBot benchmark graded 1,012 model conversations with two independent judges across six weighted dimensions. Scores help compare models under the same test, but they do not guarantee results in live conversations.
- ✓Same BrokerBot conversation framework across model choices.
- ✓Caching-aware processing estimates based on measured usage.
- ✓Explicit tradeoffs shown instead of assuming the most expensive model is best at everything.