Features
Serverless API for GLM-5.2, GLM-5.1, Qwen3.5, DeepSeek V4 Flash
Dedicated endpoints with <24h optimization turnaround
Profile-guided inference stack optimization across model/engine/kernel/hardware
152.1 tokens/s output speed for GLM-5.1 (Reasoning)
288.5 tokens/s output speed for Qwen 3.5 397B-A17B
Wafer-ai CLI for GPU-powered coding agents
GPU kernel profiling in VS Code/Cursor
KernelArena benchmark for AI-generated GPU kernels
Built-in Perfetto trace viewer and trace comparison
Cloud Compiler Analyzer (PTX/SASS) in IDE
GPU docs and remote kernel evaluation
Workspace GPU compute for coding agents
Order-of-magnitude speedups on AMD GPUs via kernel engineering
Heterogeneous compute with custom kernels (e.g., AMD MI355X)
NVFP4 quantization for Blackwell inference (e.g., Kimi K2.6)
One-command install on macOS, Linux, Windows
Run hundreds of open models locally
MLX engine for Apple Silicon (faster, less memory)
GGUF model support via llama.cpp (Ollama 0.30)
NVIDIA Nemotron 3 Ultra for high-throughput reasoning
Cloud scaling with Free, Pro, Max tiers
Run multiple cloud models in parallel (1, 3, 10)
Web-enabled cloud agents for real-time info retrieval
Fully offline operation for mission-critical work
Data never used for training; privacy-first design
CLI tool with model management and configuration
REST API for building AI applications
Upload and share private models (Pro and above)
40,000+ community integrations
Desktop app for macOS, Linux, Windows
