DeepSeek V4.1 Flash with cached input from $0.008 per 1M tokens. DeepSeek V4.1 Flash Fast at up to 322 output tokens per second.
Pantry is free and open source. The rest of the AI stack (Iron, Butter, Syrup, Whisk, Waffle) is in private early access with pricing to be determined.
Key features
Prefix caching enabled by default
1 million token context window
Chat Completions and Responses
Streaming support
Tool calls
OpenAI-compatible API
Up to 322 output tokens per second
Pantry: free, open-source vulnerability scanner for macOS and Linux
Iron: GPU kernels optimized for Apple silicon, NVIDIA CUDA, and AMD ROCm
Butter: multi-modal inference engine for text, vision and audio
Whisk: memory and context management to mix documents and context into requests
Syrup: control plane for monitoring, telemetry, access control and cost per GPU
Waffle: agent harness for building assistants
Runs on Apple silicon, NVIDIA RTX, H100, H200, B100, B200 and AMD ROCm
Local execution keeps data on your hardware by default
What makes it different
Lowest price per 1M tokens at $0.008 for cached input
Fastest inference speed at 322 output tokens per second
Built specifically for long-running agent workloads
Prefix caching on by default
Built for lower cost per token through efficiency optimization
Private by default; data stays on your hardware
Supports multiple hardware platforms with write-once kernels
Pantry offers free, open-source vulnerability scanning and patching
Enables AI on commodity hardware people already own