Today's LLM & Language Models: Fastest-Growing Projects — April 17, 2026
This week, the LLM & Language Models space saw a surge in interest around tools that enhance or optimize language model performance, as well as those that provide accessible and user-friendly interfaces for interacting with these models. We also noticed a trend towards open-source implementations of popular LLM patterns, such as Karpathy's LLM Wiki. These trends are reflected in the top-growing repositories in this space.
Arman-bd/guppylm is a ~9M parameter LLM that talks like a small fish, and has gained 2,900 stars with a growth score of 99.37. Its unique approach to generating human-like text has captured the attention of developers and researchers alike, driving its rapid growth.
Hexiecs/talk-normal is a system prompt that removes AI slop from LLM outputs, allowing them to talk like normal people. With a growth score of 83.89 and 1,102 stars, this repository is attracting interest from those looking to refine their language model's communication style.
AmitShekhariitbhu/llm-internals offers a step-by-step guide to understanding LLM internals, covering topics from tokenization to attention and inference optimization. With a growth score of 61.80 and 450 stars, this repository is becoming a go-to resource for developers looking to deepen their knowledge of language models.
Sdyckjq-lab/llm-wiki-skill is a personal knowledge base construction skill based on Karpathy's LLM Wiki method, supporting multiple platforms. Its growth score of 60.00 and 777 stars reflect the growing interest in implementing this popular pattern.
Mnfst/awesome-free-llm-apis provides a list of permanent free LLM APIs (API keys), which has attracted 2,192 stars and a growth score of 53.59. As developers continue to explore language model applications, this repository is becoming an essential resource for finding accessible APIs.
Kessler/gemma-gem runs Google's Gemma 4 model entirely on-device via WebGPU, eliminating the need for API keys or cloud services. With a growth score of 45.04 and 721 stars, this repository is gaining traction among those looking for self-contained language model solutions.
Askalf/dario is a local LLM router that allows users to access multiple providers through a single endpoint. Its growth score of 41.50 and 69 stars reflect the interest in streamlining language model interactions.
Pratiyush/llm-wiki implements Karpathy's LLM Wiki pattern, allowing users to create knowledge bases from their Claude Code, Codex CLI, Copilot, Cursor & Gemini sessions. With a growth score of 41.00 and 102 stars, this repository is part of the growing trend towards open-source LLM implementations.
Lucasastorian/llmwiki offers an open-source implementation of Karpathy's LLM Wiki pattern, allowing users to upload documents and connect their Claude account via MCP. Its growth score of 34.42 and 460 stars demonstrate the ongoing interest in this popular pattern.
Atomicmemory/llm-wiki-compiler is a knowledge compiler that takes raw sources and generates interlinked wiki outputs, inspired by Karpathy's LLM Wiki pattern. With a growth score of 33.17 and 526 stars, this repository rounds out our list of top-growing tools in the LLM & Language Models space.
Arman-bd/guppylm is a ~9M parameter LLM that talks like a small fish, and has gained 2,900 stars with a growth score of 99.37. Its unique approach to generating human-like text has captured the attention of developers and researchers alike, driving its rapid growth.
Hexiecs/talk-normal is a system prompt that removes AI slop from LLM outputs, allowing them to talk like normal people. With a growth score of 83.89 and 1,102 stars, this repository is attracting interest from those looking to refine their language model's communication style.
AmitShekhariitbhu/llm-internals offers a step-by-step guide to understanding LLM internals, covering topics from tokenization to attention and inference optimization. With a growth score of 61.80 and 450 stars, this repository is becoming a go-to resource for developers looking to deepen their knowledge of language models.
Sdyckjq-lab/llm-wiki-skill is a personal knowledge base construction skill based on Karpathy's LLM Wiki method, supporting multiple platforms. Its growth score of 60.00 and 777 stars reflect the growing interest in implementing this popular pattern.
Mnfst/awesome-free-llm-apis provides a list of permanent free LLM APIs (API keys), which has attracted 2,192 stars and a growth score of 53.59. As developers continue to explore language model applications, this repository is becoming an essential resource for finding accessible APIs.
Kessler/gemma-gem runs Google's Gemma 4 model entirely on-device via WebGPU, eliminating the need for API keys or cloud services. With a growth score of 45.04 and 721 stars, this repository is gaining traction among those looking for self-contained language model solutions.
Askalf/dario is a local LLM router that allows users to access multiple providers through a single endpoint. Its growth score of 41.50 and 69 stars reflect the interest in streamlining language model interactions.
Pratiyush/llm-wiki implements Karpathy's LLM Wiki pattern, allowing users to create knowledge bases from their Claude Code, Codex CLI, Copilot, Cursor & Gemini sessions. With a growth score of 41.00 and 102 stars, this repository is part of the growing trend towards open-source LLM implementations.
Lucasastorian/llmwiki offers an open-source implementation of Karpathy's LLM Wiki pattern, allowing users to upload documents and connect their Claude account via MCP. Its growth score of 34.42 and 460 stars demonstrate the ongoing interest in this popular pattern.
Atomicmemory/llm-wiki-compiler is a knowledge compiler that takes raw sources and generates interlinked wiki outputs, inspired by Karpathy's LLM Wiki pattern. With a growth score of 33.17 and 526 stars, this repository rounds out our list of top-growing tools in the LLM & Language Models space.