Apache TVM
by Apache Software Foundation · free · Last verified 2026-03-17
Apache TVM is an open-source machine learning compiler stack that optimizes deep learning workloads for a diverse set of hardware backends including CPUs, GPUs, FPGAs, and custom accelerators. It automates model optimization through its AutoTVM and Ansor auto-tuning systems, delivering state-of-the-art inference performance on edge targets.
https://tvm.apache.org ↗B
B—Above Average
Adoption: B+Quality: A+Freshness: B+Citations: AEngagement: F
Specifications
- License
- Apache 2.0
- Pricing
- free
- Capabilities
- model compilation, hardware-specific optimization, auto-tuning, operator fusion, quantization
- Integrations
- CUDA, Vulkan, Metal, OpenCL, LLVM, WebAssembly
- Use Cases
- edge hardware optimization, custom accelerator deployment, cross-platform model serving, research compilation targets
- API Available
- No
- SDK Languages
- python, c++
- Deployment
- edge, embedded, cloud, browser
- Rate Limits
- No limits — fully on-device
- Data Privacy
- On-device compilation and execution; no telemetry by default
- Tags
- edge, compiler, optimization, inference, hardware, deep-learning
- Added
- 2026-03-17
- Completeness
- 100%
Index Score
68Adoption
72
Quality
90
Freshness
75
Citations
85
Engagement
0