B
BaseRT by Base Compute
The fastest LLM inference runtime for Apple Silicon
Up to 6.4x faster prefill than llama.cpp33% faster decode on Qwen3Optimised for Apple M-series chipsOpen-source runtimeIntegrated with local coding agents
Opportunity score
55/100
Moderate opportunity: clear pain point but monetisation is unproven, competition is fierce, and the technical moat is replicable with enough investment.
Founder verdict
MAYBE
BaseRT proves the hunger for local AI speed, but the path to a defensible business is unclear and heavily dependent on temporary hardware advantages.
What is BaseRT by Base Compute?
BaseRT is an open-source inference runtime purpose-built for Apple Silicon (M1–M5) that delivers substantially faster LLM inference than popular alternatives like llama.cpp and Apple's own MLX, especially on prefill-bound workloads. The team from Base Compute (Melbourne & Berlin) positions it as the default backend for on-device AI and local coding agents, emphasising privacy, speed, and a one-command install. While benchmarks show impressive gains on certain models, monetisation details remain undisclosed, and the project appears to be in an early community-building phase.
Get the full BaseRT by Base Compute playbook
The complete reverse-engineering — what to copy, what to avoid, and exactly what to build instead.
- Full opportunity score across 8 dimensions
- The real problem & why customers pay
- Why it's winning — with evidence
- Complete business reverse-engineering
- Competitive advantages & moat analysis
- Weaknesses, risks & what to avoid
- Clone strategy — what to build instead
- Week-by-week MVP roadmap
- Recommended technical stack
- Founder verdict & highest-leverage move