Models
Manage real model inventory by execution path. Providers and engines are tracked separately.
Local Models
Models installed on the Mopitu IDE machine and served by the Local Engine.
| Model | Version | Status | Requests 24h | Latency | Failure Rate |
|---|---|---|---|---|---|
| Qwen 3.6 Local | 3.6 | enabled | 1,284 | 910ms | 2% |
| Kimi Local | latest | enabled | 740 | 1080ms | 3% |
| Llama 3.1 Local | 3.1 | enabled | 626 | 980ms | 1.8% |
| DeepSeek R1 Local | R1 | enabled | 514 | 1320ms | 2.5% |
| Phi 3.5 Local | 3.5 | enabled | 388 | 760ms | 1.4% |
Cloud Models
Models exposed through RegulaCore Cloud AI. RegulaCore Cloud AI is the provider, not a model.
| Model | Tier | Status | Requests 24h | Latency | Failure Rate |
|---|---|---|---|---|---|
| GPT-5 High via RegulaCore Cloud AI | Tier 1 | enabled | 2,188 | 640ms | 1% |
| Claude Sonnet via RegulaCore Cloud AI | Tier 1 | enabled | 1,742 | 690ms | 1.2% |
| DeepSeek R1 via RegulaCore Cloud AI | Tier 2 | enabled | 1,210 | 720ms | 1.6% |
| Kimi K2 via RegulaCore Cloud AI | Tier 2 | enabled | 998 | 780ms | 1.8% |
Swarm Models
Models that can be executed through MopiSwarm. MopiSwarm is the distributed execution provider, not a model.
| Model | Swarm Nodes | Status | Requests 24h | Latency | Failure Rate |
|---|---|---|---|---|---|
| Qwen 3.6 through MopiSwarm | 8 | enabled | 442 | 1180ms | 1.9% |
| Kimi through MopiSwarm | 6 | enabled | 316 | 1260ms | 2.1% |
| DeepSeek R1 through MopiSwarm | 5 | enabled | 284 | 1480ms | 2.7% |
| Llama 3.1 through MopiSwarm | 7 | disabled | 76 | 1090ms | 2.3% |