Maple
Privacy-first LLM inference with AMD SEV-SNP
Maple delivers privacy-preserving AI inference through AMD SEV-SNP confidential VMs. Accessed via a local proxy that routes requests through verified confidential compute nodes, ensuring prompts never leave the encrypted enclave. Designed for teams that need strong data residency and auditability.
API Features
OpenAI Compatible ✓
Streaming ✓
Function Calling —
Batch Inference —
Vision ✓
Embeddings —
Models & Pricing
| Model | Family | Context | Max Out | Input /M | Output /M | Free |
|---|---|---|---|---|---|---|
| GPT-OSS 120B | OpenAI GPT | 128K | — | $1.50 | $2.50 | — |
| Kimi K2.5 | Moonshot AI | 256K | — | $3.00 | $10.50 | — |
| DeepSeek R1 0528 | DeepSeek | 128K | — | $3.00 | $10.50 | — |
| Llama 3.3 70B | Meta Llama | 128K | — | $3.50 | $5.50 | — |
| Qwen3-VL 30B | Qwen | 256K | — | $2.50 | $8.00 | — |