Loading leaderboard...
Community benchmark results from Anubis OSS
Real-world LLM performance on Apple Silicon - tokens/sec, power draw, TTFT, and more. Run a benchmark in Anubis and upload your results to see how your Mac stacks up.
Loading leaderboard...
Note: When this data is collected and submitted from Apple Computers, only systems submitting runs from an Ollama backend contain the full gamut of datapoints - LM Studio, MLX and others do not expose all performance metrics, such as prompt evaluation time. But the dataset is robust, even without Ollama as the runtime. Enjoy! -JT