| Company | ARR | Growth | Customers | Notes |
|---|---|---|---|---|
| CloudNC | $6.3M | 5.5× | 915 | 85% inbound · 5-day sales cycle · 1:1 GTM efficiency |
| Lambda Function | <$100K | n/a | Handful | Pre-PMF · Supervised ML |
| Toolpath | Pre-PMF | n/a | Low double digits | Pre-PMF · CTO departed March 2026 · 3-axis ceiling |
| LimitlessCNC | Pre-PMF | n/a | ~3–5 enterprises | Pre-PMF · 3–6 month on-site deployments · high-volume only |
| up2parts (Sandvik) | Pre-PMF | n/a | ~0 | Pre-PMF · Sandvik-backed · no meaningful traction |
| Manukai (Siemens) | Pre-PMF | n/a | ~0 | Pre-PMF · Siemens Xcelerator-backed · no meaningful traction |
Three fundamentally different architectural approaches exist in AI CAM. The approach determines the capability ceiling, not just current performance.
A critical structural point: every other company in this space relies on ModuleWorks or Siemens toolpath kernels for their underlying geometry engine. This creates a hard ceiling on what they can achieve. CloudNC's toolpath engine is built entirely in-house, giving it full control over the technology stack and no dependency on third-party kernels.
| Dimension | Physics-Based AI CloudNC |
Supervised ML LimitlessCNC · Lambda · up2parts · Manukai |
Game-Playing AI Toolpath |
|---|---|---|---|
| Core method | Purpose-built CAM kernel with physics-based solvers and 11 proprietary AI primitives | Pattern-matching on historical machining data | Reinforcement learning over strategy space |
| Novel parts | ✓ Works on unseen parts from first principles | ✗ Degrades or fails on unseen geometry | ~ Can generalise in theory; struggles in practice |
| Complexity | Full 3+2 axis, whole-part strategies | 3-axis ceiling, feature-based | Stuck at 3-axis |
| Deployment | Cloud-native SaaS; full value in days | On-premise; 3–9 months before value | Web + Fusion plugin |
| Time-to-value | Days (full capability) | Months (data ingestion required) | Days (limited capability) |
| Scale economics | Zero marginal cost: same engine for all | Linear: each customer needs dedicated training | Product-led but ceiling-limited |
CloudNC has invested 10 years and $70M+ building its physics-based CAM kernel. This is not a model you retrain — it is a purpose-built computational geometry engine with 11+ proprietary AI primitives spanning feature recognition, strategy solving, collision avoidance, and toolpath optimization.
The companies with the deepest domain expertise in manufacturing software concluded that building AI CAM from scratch was not viable on any reasonable timescale.
Machine learning trained on historical machining programs can only recommend strategies similar to what it has seen. Manufacturing's long tail — novel geometries, unusual materials, non-standard setups — is precisely where automation creates the most value, and precisely where supervised ML fails. Physics-based reasoning generalises to any geometry because the underlying physics doesn't change.
Fusion 360 is a cloud-native platform serving hundreds of thousands of users making everything from prototypes to production parts. Any AI CAM partner must meet five non-negotiable requirements to integrate with Fusion's model.
| Requirement | CloudNC | Limitless | Lambda | Toolpath | up2parts | Manukai |
|---|---|---|---|---|---|---|
| Cloud-native Fusion is a cloud platform |
✓ | ✗ | ~ | ✓ | ~ | ~ |
| Instant integration Not 3–6 month on-site deployment |
✓ | ✗ | ~ | ✓ | ✗ | ✗ |
| Works on novel parts Fusion users make everything |
✓ | ✗ | ✗ | ~ | ✗ | ✗ |
| Scales to 100K+ users Platform-level distribution |
✓ | ✗ | ✗ | ~ | ✗ | ✗ |
| Remote deployment No on-site required |
✓ | ✗ | ✓ | ✓ | ~ | ~ |
CloudNC is expanding beyond CAM automation. The capabilities shipping over the next 3–9 months contain everything competitors offer as a subset.
Instant machining strategies for any geometry. 915 factories. $6.3M ARR.
AI-generated CNC quotes from 3D geometry. Cycle times, costs, feasibility — in seconds.
Design-for-manufacturability feedback before the part reaches the shop floor.
Fully autonomous CNC programs from raw 3D geometry. Zero human input.
| Capability | CloudNC | Limitless | Lambda | Toolpath | up2parts | Manukai |
|---|---|---|---|---|---|---|
| CAM Automation | ✓ | ~ | ~ | ~ | ~ | ~ |
| Multi-axis (3+2) | ✓ | ✗ | ✗ | ✗ | ✗ | ✗ |
| Novel geometry | ✓ | ✗ | ✗ | ~ | ✗ | ✗ |
| Agentic quoting | Soon | ✗ | ✗ | ✗ | ~ | ✗ |
| Agentic DFM | Soon | ✗ | ✗ | ✗ | ✗ | ✗ |
| Autonomous CAM | ~6mo | ✗ | ✗ | ✗ | ✗ | ✗ |
Every capability CloudNC adds expands the value it delivers through Fusion. No competitor is building toward this breadth.