AI Cloud · pricing
AI Cloud pricing
One seat price covers the workspace, two open-weight models and the security controls. Add-ons price support levels, compliance evidence and extra capacity.
Plans
annual plan
Billed annually, by card or invoice.
monthly plan
Billed monthly, by card.
the seat
What the seat includes
- Single sign-on through SAML or OIDC, SCIM provisioning, roles and audit logs, all kept in your account
- Controls documentation aligned to SOC 2
- Zero data retention at Clover Tech, and conversation retention set by your admins
- Spend controls: per-user metering, monthly and daily caps and a per-agent kill switch in development
- The workspace, with projects that switch between models per chat
- Agents with instructions, file search and code execution
- Three RAG connections, each to one internal data source
- Two open-weight models, plus Claude or GPT on your own vendor key with no Clover Tech fee
- Automatic routing of routine work to open-weight models and harder tasks to frontier models in development
- Standard support, covering installation, production issues and model updates
- A usage dashboard that reports by team, with a minimum group size in development
cloud compute
Compute is on your own cloud bill
Your cloud provider bills GPUs, storage and network to your account, and frontier model usage goes on your own vendor or cloud bill.
For teams under 50 users, compute costs as much as the seats or more. The example bills estimate compute for each configuration.
add-ons
Add-ons
| add-on | what it covers | price |
|---|---|---|
| products | ||
| Clover Code | VS Code in the browser, connected to an open-weight coding model in your account | +$5 per developer a month, coding model included |
| MCP server | Lets an AI tool deploy and run Clover Tech Marketplace apps in your account | $1,000 a year; included with agent payments |
| Agent payments in development | USDC on Base and Solana, Bitcoin over Lightning, a dedicated BTCPay node | $1 per 1,000 stablecoin payments, plus $6 a month per BTCPay node |
| support | ||
| Standard | Email, 8 AM to 8 PM ET (5 AM to 5 PM PT), Monday to Friday; production issues answered the next business day | Included |
| Priority | Production issues answered within 4 business hours | +$2 per seat a month, $2,000 a year minimum |
| 24 × 7 | Production issues answered within 1 hour, around the clock | +$3 per seat a month, $10,000 a year minimum |
| compliance evidence | ||
| Security report | Quarterly deployment report and annual SOC 2 evidence pack, built in your account from audit logs | $1,200 a year |
| Compliance mapping | Controls documentation for one framework beyond SOC 2, starting with GDPR | $15,000 a year per framework |
| Model evaluation report in development | Cost per session for the same users before and after a model change | $5,000 a year |
| capacity | ||
| RAG connection | Each internal data source beyond the three included | $1,000 a year |
| Open-weight model | Each model beyond the two included. A new release of an included model runs beside it on trial at no charge. | $1,000 a year |
| people | ||
| Integration and workflow expert | Integrations and workflow design, which support does not cover | $12,000 a year |
Per deployment
RAG connections, extra models, compliance mapping, the security report and the model evaluation report apply per deployment, so a second deployment, such as one in an EU region, repeats them. Add-ons bill on the same cycle as seats.
Every support level
Covers installation, production issues, model updates and replay of error logs, plus log forwarding to your MSSP. Your cloud provider runs the compute and network, so Clover Tech commits to response times and offers no uptime SLA. End-user support stays with your staff.
examples
Example annual bills
Five configurations on annual billing, with the compute each would need on AWS.
| buyer | seats | add-ons | Clover Tech per year | estimated compute per year |
|---|---|---|---|---|
| Family office, 12 users | $2,880 | Security report $1,200 | $4,080 | $12,300–$16,3001 L40S |
| InfoSec team, 20 users | $4,800 | 24 × 7 support $10,000 (minimum), 3 extra RAG connections $3,000, 1 extra open-weight model $1,000, security report $1,200 | $20,000 | $24,700–$32,6002 L40S |
| Trading team, 25 users | $6,000 | Agent payments: 240,000 payments $240, one BTCPay node $72; MCP server included | $6,312 | $12,300–$16,3001 L40S |
| DevOps team, 40 users | $9,600 | Clover Code for 40 developers $2,400, Priority support $2,000 (minimum), 2 extra RAG connections $2,000 | $16,000 | $24,700–$32,6002 L40S |
| Public company, 3,000 users | $648,000after 10% off | 24 × 7 support $108,000, integration and workflow expert $12,000, compliance mapping for GDPR $15,000, security report $1,200, model evaluation report $5,000, 5 extra RAG connections $5,000, Clover Code for 300 developers $18,000 | $812,200 | About $482,0008 H100 |
Counts and add-ons are examples, and seats are at the annual rate. Compute estimates use AWS us-east-1 GPU prices, from a one-year savings plan to on-demand, sized for two Qwen3.6 models with 1 in 10 to 1 in 20 users active at once: one NVIDIA L40S for up to about 50 users, two for coding or three-model teams, eight H100s for 3,000 users. A load test on your own usage sets the final size. Frontier model usage and blockchain network fees are billed separately.
build or buy
What running it yourself costs
Running the same open-source stack yourself adds an estimated $1,500 to $2,500 a month of operations work (Navya AI, Cost to host a private LLM). That is the work the maintained install takes on: patches, tested upgrades and controls documentation.
next step
Request a quote for your firm
Send your seat count, your cloud provider and the add-ons you are weighing. We reply by email with a quote.