Quick facts
- Best for
- whichllm local LLM picker for your actual AI hardware
- Pricing
- Free
- Editor rating
- 4.5 / 5
- Community saves
- 0
About whichllm
whichllm is an open-source command-line tool that recommends local LLMs for a user’s actual hardware. It ranks models by VRAM fit, speed, and benchmark quality using recency-aware model data, supports GPU simulation, markdown and JSON output, and can generate run commands or Python snippets for selected models.
Pros
- Ranks local LLMs by VRAM fit, expected speed, and benchmark quality
- Simulates GPUs such as RTX 4090, multi-GPU workstations, or custom VRAM limits
- Supports one-off uvx use plus uv, Homebrew, and pip installation paths
- Outputs Markdown or JSON for scripts, Slack, Discord, and documentation
- Includes planning commands to find the GPU needed for a target model
Cons
Pricing
Open-source project
$0
- • Source code is available in the public repository
- • Review the repository license and setup notes before production use
- • No hosted SaaS pricing was found in the official source