Beating Pokemon Red with RL and <10M Parameters
Hi everyone! After spending hundreds of hours, we're excited to finally share our progress in developing a reinforcement learning system to beat Pokémon Red. Our system successfully completes the game using a policy under 10M parameters, PPO, and a few novel techniques. With the release of Claude Plays Pokémon, now feels like the perfect time to showcase our work. We'd love to get feedback!
Beating Pokemon Red with RL and <10M Parameters launched on March 5, 2025 with 183 votes, #213 of 7,215 launches that month and more than 95% of that year's launches. The site was still online in September 2026, 1 year on.
In plain words
A reinforcement learning system that completes Pokémon Red using a policy under 10 million parameters and the PPO algorithm. The project demonstrates novel techniques for training game-playing AI agents on complex, turn-based RPGs. Designed for developers and AI enthusiasts interested in reinforcement learning applications, the work showcases how relatively compact neural networks can solve full-length video games through iterative training and optimization.
written from the facts on this page · September 2026
From the sources
Hi! Since 2020, we’ve been developing a reinforcement learning (RL) agent to beat the 1996 game Pokémon Red. As of February 2025, we are able to beat Pokémon Red with Reinforcement Learning using a <10 million parameter policy (60500x smaller than DeepSeekV3) and with minimal simplifications. The output is not a policy capable of beating Pokémon, but a technique for producing solutions to Pokémon.
Hi! Since 2020, we’ve been developing a reinforcement learning (RL) agent to beat the 1996 game Pokémon Red. As of February 2025, we are able to beat Pokémon Red with Reinforcement Learning using a <10 million parameter policy (60500x smaller than DeepSeekV3) and with minimal simplifications. The output is not a policy capable of beating Pokémon, but a technique for producing solutions to Pokémon. This website describes the system’s current state. All code is open sourced and available for you, the reader, to try . If you want to contact us, please join our Discord server (linked below) or email us at pokemonreinforcementlearning <at> gmail <dot> com . As improvements to the…from drubinstein.github.io
Does a similar job
all alternatives →
Jev Plays Pokémon Red27d ago · jev-pokemon.vercel.app · ▲282Watch the AI decision model Jev play Pokémon Red in its entirety, live.
Watch a neural net learn to play SnakeMay 2026 · ppo.gradexp.xyz · ▲205In browser PPO training demo, made possible by tinygrad: TinyJit -> WebGPU kernels. Requires WebGPU.
Continue? Y/N: A 60-second game about AI agent permission fatigueMay 2026 · llmgame.scalex.dev · ▲386A 60-second game about LLM permission fatigue. Claude Code needs your approval — but are you really reading the commands?
Play rock paper and scissors against a untrained neural network2019 · github.com · ▲101Can a neural network predict your next move on a game of rock, paper and scissor? - victorqribeiro/jokenpo
LLM plays Pokémon (open sourced)2025 · github.com · ▲186I built a bot that plays Pokémon FireRed. It can explore, battle, and respond to game events. Farthest I made it was Viridian Forest. I paused development a couple months ago, but given the launch of ClaudePlaysPokemon, decided to open source!
FabraixMay 2026 · fabraix.com · ▲196We built Adversarial Cost to Exploit (ACE), a benchmark that measures the token expenditure an autonomous adversary must invest to breach an LLM agent. Instead of binary pass/fail, ACE quantifies adversarial effort in dollars, enabling game-theoretic analysis of when an attack is economically rational. We tested six budget-tier models (Gemini Flash-Lite, DeepSeek v3.2, Mistral Small 4, Grok 4.1 Fast, GPT-5.4 Nano, Claude Haiku 4.5) with identical agent configs and an autonomous red-teaming attacker. Haiku 4.5 was an order of magnitude harder to break than every other model; $10.21 mean…
More life & fun this month
the category →
A free city-building game in your browser. You are the mayor. You know what the city needs. Now get it past the zoning board, the neighbors, the appeals court and the next election.
Life & fun · 2d ago · permit.petercai.com
- IV
Life & fun · 14d ago · tanzaniaeducationproject.org
Turn a phone, tablet or TV into a sign. Type a message and it fills the screen: welcome signs, timers, Wi-Fi passwords and more. No app, no account, free.
Life & fun · 15d ago · bigwords.page
Play a free virtual guitar online with your keyboard or by touch. Slow songs down and see every chord, string and strum. Acoustic, classic, electric and bass.
Life & fun · 9d ago · guitoro.com
Carrier settings from iPhone, Pixel and Galaxy firmware, decoded and compared: APNs, VoLTE, 5G and Wi-Fi Calling per carrier, and what each build changed.
Life & fun · 13d ago · carrierexplode.com
Life & fun · 23d ago · space.bl2.net
Source: Show HN post ↗
Launched alongside, March 2025
the whole month →
Mimic Human Research & Save Findings in AI Knowledge Base
AI · 2025 · sider.ai
