Look up AI model specs, benchmarks, parameters, pricing, and memory guidance.
npx skills add https://github.com/wenerme/ai --skill model-reference
Please help me install this Agent Skill. Skill: model-reference Source: https://github.com/wenerme/ai/tree/main/skills/model-reference Command: npx skills add https://github.com/wenerme/ai --skill model-reference
This Skill removes the guesswork in choosing AI models by giving you a quick, structured reference for capabilities, benchmarks, recommended sampling settings, memory needs, and pricing across major LLM/VLM and generation families.
Ask an AI: "Use the model-reference skill to recommend sampling parameters for a reasoning task and estimate memory needs for a selected model family, then point me to the relevant references file for more details."
High-intent search queries and answers about installing and using this skill.
To find the right AI model for coding or reasoning, compare model families using quick specs, benchmark signals, and recommended task modes. This provides structured guidance to reduce trial-and-error when selecting a model.
Recommended sampling parameters for AI models include suggested temperature, top_p, and top_k settings tailored for specific task modes like reasoning, coding, or creative generation, which you can look up in family-specific reference files.
Estimate pricing and memory requirements for AI model deployment by consulting a reference overview that provides cost details and VRAM/precision guidance across major LLM and VLM families.
Yes, you can compare model characteristics across families like Qwen, DeepSeek, and Llama by using overview and topic reference files to evaluate capabilities, benchmark performance, and common configuration patterns.
The best way to set temperature and top_p for AI generation tasks is to use provided parameter tables that offer practical guidance and recommended settings based on the specific model family and task mode.
Yes, the model reference provides VRAM sizing guidance by outlining memory requirements across different precision levels, allowing you to plan deployment resources effectively for your selected model family.
agentic-inSelect inference paths, serving stacks, and benchmarking plans for ML models.
letta-aiRetrieve and compare top model standings from MTEB and embedding benchmark leaderboards.
Prathmesh2000Analyze and estimate token usage for LLM API calls.
jdforsytheDetect and report performance inefficiencies using static analysis, benchmarking, and profiling.
