Pricing
CLI
FreeDesign on your own, on your own machine.
Get startedIncludes:
- The full agent: do, create, init, check, watch
- Bring your own Claude or GPT-5 key
- Runs on your KiCad files, in your own git repo
- Plain markdown, JSON and KiCad output. No lock-in.
- Runs locally, never metered
- Community support
- Most popular
Cloud
$49per user / monthHosted runs on private repos, with a web viewer.
Let's talkEverything in CLI, plus:
- Hosted runs, no local setup, from anywhere
- Private repositories
- Web viewer: chat, live schematic and board render, ERC/DRC status
- Run history and shareable check reports
- One-click gerber, DXF/STEP, render and BOM export
- BYO key, or managed inference with 200 credits a seat
Team
$49per user / monthCloud, with governance and CI for a whole team.
Let's talkEverything in Cloud, plus:
- CI bot: run check as a required PR status check
- Shared, versioned constraint libraries across the org
- SSO / SAML, role management, central billing
- Managed credits pooled across the team
- Shared run history
Enterprise
CustomannualSelf-hosted, one flat price a year.
Talk to salesEverything in Team, plus:
- Self-hosted or VPC, so design data never leaves your network
- Altium support beyond KiCad
- RBAC, security review, dedicated support and an SLA
- An audit trail your compliance team can hand over
- Flat annual license, no per-run metering
Three ways to pay for a run.
Every design run is an agent loop that spends model tokens. Copperhead never subsidizes those tokens. You either bring your own key, or run on managed inference that is metered above cost.
- Mode A
Bring your own key
Supply your own Claude or GPT-5 key. Copperhead runs the agent but never touches your token bill. You pay for the platform and your provider bills the tokens. Available on every paid tier.
- Mode B
Managed inference
Let Copperhead run the model. Each seat includes a monthly allowance of design credits; overage is metered at model cost + 20%, or bought as top-up packs. Hard spend caps on every seat and org, with runaway bills refunded.
- Mode C
Flat annual
No per-run metering. A flat annual license sized to expected volume, with inference self-hosted in your own environment or routed through a private endpoint. One figure a year for procurement.
How the four tiers compare.
| Feature | CLI | Cloud | Team | Enterprise |
|---|---|---|---|---|
| Price | Free | $49 / user / mo | $49 / user / mo + $199 / mo | Custom, annual |
| Repos | Your own, local | Private, hosted | Private, hosted | Private / VPC / self-hosted |
| Inference | BYO key | BYO key or managed | BYO key or managed (pooled) | Self-hosted / private, flat |
| Managed credits | n/a | 200 / seat / mo | 200 / seat, pooled | Negotiated volume |
| Overage | n/a | model cost + 20% | model cost + 20% | None (flat) |
| CI bot (check) | Manual | Manual | Required PR check | Required PR check |
| Shared constraint libraries | Local | Per-user | Org-wide | Org-wide |
| SSO / RBAC | n/a | n/a | SSO | SSO + RBAC |
| Audit trail artifact | In repo | In repo | In repo | Compliance-grade |
| Support | Community | Priority | Dedicated + SLA |
The free CLI stays free.
Paid plans only ever add hosting, collaboration and governance on top of files you already own. They never wall off the work itself.
Building in the open? Copperhead Cloud is free for public repositories released under CERN-OHL-S or an OSHWA-certified license. Managed inference is included within fair-use limits.
Startups and non-profits: 50% off · accredited education: free
Plain text, in your repo, at every tier
Paid plans add hosting, collaboration and governance on top of files you fully own. No proprietary format, no lock-in.
Cancel any time
Drop a subscription and you fall back to the free CLI, working on the same files.
A grant-funded commons
The free CLI is funded by grants rather than by hosted revenue, so its funding does not depend on the paid tiers.
Questions on volume, credits or self-hosting? Talk to us or book a demo call.