Today's AI Research: Fastest-Growing Projects — May 01, 2026
Today's AI Research, we saw significant growth in tools focused on multimodal intelligence, autonomous agents, and large language model (LLM) research. The trend suggests a shift towards more complex and interactive AI systems that can perceive, reason, and act in various environments.
fkyah3/opencode-yg is a research fork of opencode demonstrating Language Anchoring, which makes LLMs think consistently in a specific language. With a growth score of 35.33 and 27 stars, this repository is gaining attention for its potential to improve the consistency and reliability of LLMs.
AutoMedBench/AutoMedBench is a Medical AutoResearch Benchmark for Autonomous AI Agents, providing a standardized evaluation framework for medical AI research. Its growth score of 16.39 and 22 stars indicate increasing interest in applying autonomous agents to healthcare applications.
matrix-agent/awesome-agentic-world-modeling is a curated collection of resources on Agentic World Modeling, including papers, code, and datasets. With a growth score of 15.29 and an impressive 142 stars, this repository serves as a valuable resource for researchers exploring the foundations and capabilities of agentic world modeling.
Hedlen/Awesome-Multimodal-Intelligence is another curated collection focused on multimodal intelligence research, covering VLMs, VLAs, World Models, and embodied AI. Its growth score of 9.50 and 32 stars demonstrate growing interest in next-generation intelligent agents that can perceive and interact with their environment.
thunlp/OPD is the official repository for a paper on "Rethinking On-Policy Distillation of Large Language Models," exploring new approaches to distilling large language models. With a growth score of 7.97 and 192 stars, this repository highlights the ongoing effort to improve LLMs' efficiency and effectiveness.
gameworld-project/gameworld is working towards standardized and verifiable evaluation of multimodal game agents, aiming to create a common framework for testing AI agents in game environments. Its growth score of 7.78 and 165 stars indicate interest in applying multimodal intelligence to gaming and simulation applications.
7WaySecurity/ai_osint provides curated resources for AI OSINT (Open-Source Intelligence), including techniques and tools for discovering exposed LLM endpoints, leaked AI API keys, and misconfigured vector databases. With a growth score of 7.70 and 72 stars, this repository highlights the growing importance of security in AI research.
kokolerk/TCOD explores Temporal Curriculum in On-Policy Distillation for Multi-turn Autonomous Agents, aiming to improve the efficiency and effectiveness of autonomous agents. Its growth score of 7.50 and 24 stars demonstrate interest in optimizing agent performance through novel distillation methods.
earleensarellano35823414097/WorpGPT-Latest-2026-AllPrompts offers a comprehensive Red Teaming framework for testing LLM robustness against adversarial prompt engineering and jailbreak vectors. With a growth score of 7.25 and 40 stars, this repository underscores the need for robustness testing in large language models.
zubair-trabzada/ai-trading-claude is an AI trading research engine for Claude Code, analyzing stocks, options strategies, sector rotation, and portfolio analysis through various skills and parallel agents. Its growth score of 5.81 and 90 stars demonstrate interest in applying AI to finance and trading applications.
fkyah3/opencode-yg is a research fork of opencode demonstrating Language Anchoring, which makes LLMs think consistently in a specific language. With a growth score of 35.33 and 27 stars, this repository is gaining attention for its potential to improve the consistency and reliability of LLMs.
AutoMedBench/AutoMedBench is a Medical AutoResearch Benchmark for Autonomous AI Agents, providing a standardized evaluation framework for medical AI research. Its growth score of 16.39 and 22 stars indicate increasing interest in applying autonomous agents to healthcare applications.
matrix-agent/awesome-agentic-world-modeling is a curated collection of resources on Agentic World Modeling, including papers, code, and datasets. With a growth score of 15.29 and an impressive 142 stars, this repository serves as a valuable resource for researchers exploring the foundations and capabilities of agentic world modeling.
Hedlen/Awesome-Multimodal-Intelligence is another curated collection focused on multimodal intelligence research, covering VLMs, VLAs, World Models, and embodied AI. Its growth score of 9.50 and 32 stars demonstrate growing interest in next-generation intelligent agents that can perceive and interact with their environment.
thunlp/OPD is the official repository for a paper on "Rethinking On-Policy Distillation of Large Language Models," exploring new approaches to distilling large language models. With a growth score of 7.97 and 192 stars, this repository highlights the ongoing effort to improve LLMs' efficiency and effectiveness.
gameworld-project/gameworld is working towards standardized and verifiable evaluation of multimodal game agents, aiming to create a common framework for testing AI agents in game environments. Its growth score of 7.78 and 165 stars indicate interest in applying multimodal intelligence to gaming and simulation applications.
7WaySecurity/ai_osint provides curated resources for AI OSINT (Open-Source Intelligence), including techniques and tools for discovering exposed LLM endpoints, leaked AI API keys, and misconfigured vector databases. With a growth score of 7.70 and 72 stars, this repository highlights the growing importance of security in AI research.
kokolerk/TCOD explores Temporal Curriculum in On-Policy Distillation for Multi-turn Autonomous Agents, aiming to improve the efficiency and effectiveness of autonomous agents. Its growth score of 7.50 and 24 stars demonstrate interest in optimizing agent performance through novel distillation methods.
earleensarellano35823414097/WorpGPT-Latest-2026-AllPrompts offers a comprehensive Red Teaming framework for testing LLM robustness against adversarial prompt engineering and jailbreak vectors. With a growth score of 7.25 and 40 stars, this repository underscores the need for robustness testing in large language models.
zubair-trabzada/ai-trading-claude is an AI trading research engine for Claude Code, analyzing stocks, options strategies, sector rotation, and portfolio analysis through various skills and parallel agents. Its growth score of 5.81 and 90 stars demonstrate interest in applying AI to finance and trading applications.