Abstract

We introduce Voyager, the first LLM-powered embodied lifelong learning agent in Minecraft that continuously explores the world, acquires diverse skills, and makes novel discoveries without human intervention. Voyager consists of three key components: 1) an automatic curriculum that maximizes exploration, 2) an ever-growing skill library of executable code for storing and retrieving complex behaviors, and 3) a new iterative prompting mechanism that incorporates environment feedback, execution errors, and self-verification for program improvement. Voyager interacts with GPT-4 via blackbox queries, which bypasses the need for model parameter fine-tuning. The skills developed by Voyager are temporally extended, interpretable, and compositional, which compounds the agent's abilities rapidly and alleviates catastrophic forgetting. Empirically, Voyager shows strong in-context lifelong learning capability and exhibits exceptional proficiency in playing Minecraft. It obtains 3.3x more unique items, travels 2.3x longer distances, and unlocks key tech tree milestones up to 15.3x faster than prior SOTA. Voyager is able to utilize the learned skill library in a new Minecraft world to solve novel tasks from scratch, while other techniques struggle to generalize.

Voyager discovers new Minecraft items and skills continually by self-driven exploration, significantly outperforming the baselines.

Introduction

Building generally capable embodied agents that continuously explore, plan, and develop new skills in open-ended worlds is a grand challenge for the AI community. Classical approaches employ reinforcement learning (RL) and imitation learning that operate on primitive actions, which could be challenging for systematic exploration, interpretability, and generalization. Recent advances in large language model (LLM) based agents harness the world knowledge encapsulated in pre-trained LLMs to generate consistent action plans or executable policies. They are applied to embodied tasks like games and robotics, as well as NLP tasks without embodiment. However, these agents are not lifelong learners that can progressively acquire, update, accumulate, and transfer knowledge over extended time spans.

Let us consider Minecraft as an example. Unlike most other games studied in AI, Minecraft does not impose a predefined end goal or a fixed storyline but rather provides a unique playground with endless possibilities. An effective lifelong learning agent should have similar capabilities as human players: (1) propose suitable tasks based on its current skill level and world state, e.g., learn to harvest sand and cactus before iron if it finds itself in a desert rather than a forest; (2) refine skills based on environment feedback and commit mastered skills to memory for future reuse in similar situations (e.g. fighting zombies is similar to fighting spiders); (3) continually explore the world and seek out new tasks in a self-driven manner.

Voyager Components

We introduce Voyager, the first LLM-powered embodied lifelong learning agent to drive exploration, master a wide range of skills, and make new discoveries continually without human intervention in Minecraft. Voyager is made possible through three key modules: 1) an automatic curriculum that maximizes exploration; 2) a skill library for storing and retrieving complex behaviors; and 3) a new iterative prompting mechanism that generates executable code for embodied control. We opt to use code as the action space instead of low-level motor commands because programs can naturally represent temporally extended and compositional actions, which are essential for many long-horizon tasks in Minecraft. Voyager interacts with a blackbox LLM (GPT-4) through prompting and in-context learning. Our approach bypasses the need for model parameter access and explicit gradient-based training or finetuning.

Voyager consists of three key components: an automatic curriculum for open-ended exploration, a skill library for increasingly complex behaviors, and an iterative prompting mechanism that uses code as action space.

Automatic Curriculum

Automatic curriculum. The automatic curriculum takes into account the exploration progress and the agent's state to maximize exploration. The curriculum is generated by GPT-4 based on the overarching goal of "discovering as many diverse things as possible". This approach can be perceived as an in-context form of novelty search.

Skill Library

Skill library. Top: Adding a new skill. Each skill is indexed by the embedding of its description, which can be retrieved in similar situations in the future. Bottom: Skill retrieval. When faced with a new task proposed by the automatic curriculum, we perform querying to identify the top-5 relevant skills. Complex skills can be synthesized by composing simpler programs, which compounds Voyager's capabilities rapidly over time and alleviates catastrophic forgetting.

Iterative Prompting Mechanism

Left: Environment feedback. GPT-4 realizes it needs 2 more planks before crafting sticks. Right: Execution error. GPT-4 realizes it should craft a wooden axe instead of an acacia axe since there is no acacia axe in Minecraft.

Self-verification. By providing the agent's current state and the task to GPT-4, we ask it to act as a critic and inform us whether the program achieves the task. In addition, if the task fails, it provides a critique by suggesting how to complete the task.

Film Action Doble Farsi 2025 Review

By 2025, the "Film Action Double Farsi" market has proven resilient and adaptable. It has transitioned from a low-fidelity translation service to a high-tech, culturally distinct industry. While it undermines intellectual property laws and facilitates state-sanctioned censorship, it also fulfills a vital cultural function: it acts as a window to the world for a population living under strict information control. The "Double Farsi" action film is no longer just a pirated copy; it is

Searching for 2025 action movies dubbed in Farsi (Persian) reveals several highly anticipated global releases expected to hit streaming platforms with "Doble Farsi" options.

While the 2025 release calendar is still evolving, here are some of the biggest action titles that Persian-language platforms like Farsinama, Filimo, and Namava typically feature: Top Anticipated 2025 Action Releases

Black Bag: A high-stakes espionage thriller that has already garnered significant critical buzz .

Predator: Badlands: The latest entry in the iconic franchise, focusing on a more grounded and intense survival-action experience .

One Battle After Another: An epic-scale combat film that ranks high on most "must-watch" lists for 2025 .

Sisu: Road to Revenge: A sequel to the 2022 cult hit, continuing the story of the near-immortal lone commando .

Deep Cover: An undercover crime-action thriller expected to be a major streaming hit .

Superman: The reboot of the DC flagship hero, which will undoubtedly receive professional Farsi dubbing for international markets . Where to Find "Doble Farsi" Content

To watch these and other 2025 action films with Persian dubbing, you can use specialized platforms: film action doble farsi 2025

Filimo & Namava: The largest domestic Iranian platforms, which usually provide professional dubbing within days of a film's digital release.

Farsinama: A premier source for Persian language entertainment available worldwide .

YouTube Channels: Look for channels like Epoka e re or dedicated Farsi-dubbing teams that upload trailers and full films. If you'd like, I can: Help you find official release dates for specific titles.

Recommend the highest-rated action films currently available in Farsi.

Check which of these movies are available on specific streaming platforms you use. Let me know how you'd like to narrow down your search. Best New Action Movies of 2025 (Avatar: Fire and Ash)

Whether you are looking for high-octane blockbusters or gritty spy thrillers, 2025 is shaping up to be a massive year for action cinema in the Persian-speaking world. Local streaming platforms and dubbing studios have already begun releasing "Doble Farsi" (Persian dubbed) versions of some of the year's most anticipated titles.

Here are the top action movies of 2025 available with Persian dubbing: Must-Watch Global Blockbusters The Fantastic Four: First Steps

The landscape for Farsi-dubbed action films in 2025 features a mix of high-octane Hollywood sequels and specialized international releases. This report covers the most anticipated and currently available titles for Persian-speaking audiences. Top Anticipated & Recent Releases (2025)

The following films have been identified as the leading action entries for 2025 with Persian dubbing either planned or already available on platforms like Aparat and Digikala Mag: Sisu: Road to Revenge By 2025, the "Film Action Double Farsi" market

(2025): A highly violent sequel following the "man who wouldn't die" as he seeks bloody revenge for his family. This film is ranked as a top action choice for its elevated production value and the addition of Stephen Lang as the antagonist. Mission: Impossible – Dead Reckoning Part Two

(2025): The final installment of the iconic franchise featuring Tom Cruise, widely expected to be a primary focus for Persian dubbing studios due to its global popularity. Mad Max: The Wasteland

(2025): Directed by George Miller and starring Tom Hardy, this sequel to Fury Road is one of the most anticipated blockbusters of the year.

(2025): An expansion of the John Wick universe, bringing stylized, high-stakes combat to the forefront.

(2025): A martial-arts focused action film currently trending on Persian streaming sites with an "exclusive dub". The Things You Kill

(2025): A gritty mystery-action hybrid about a university professor caught in a web of family secrets and revenge. Dubbing and Distribution Channels

For viewers looking for these titles, the most reliable sources for "Doble Farsi" content include:

Aparat: Often hosts community-uploaded and "exclusive" Persian dubs of new international releases.

Digital Entertainment Magazines: Sites like Fararu and Digikala Mag provide updated rankings and release schedules specifically for the Iranian market. The "Double Farsi" action film is no longer

YouTube: Channels dedicated to Persian cinema frequently upload high-quality dubbed versions and reviews of 2025 action hits.

۱۲ فیلم اکشن برتر سال ۲۰۲۵ - فرارو

The most intriguing word is "doble." In Spanish, it means double; in English phonetic slang, it evokes "double action" (a revolver trigger mechanism). For the Iranian viewer searching in Latin script (Finglish), this is likely a hybrid error for "dobleh" (dubbed) or a desire for "double the length/action."

This misspelling is crucial. It suggests a viewer navigating a restrictive digital landscape. Official streaming platforms in Iran (Filimo, Namava) are censored, while VPNs lead to international sites. The user is frantically typing a broken code—"action doble farsi"—hoping an algorithm will understand: Give me a film where Iranians fight back, with twice the explosions, and no subtitles.

By: The Cinematic Forecast Team

As we approach the midpoint of the decade, the global appetite for high-octane cinema shows no signs of slowing down. For the Persian-speaking community worldwide—from Tehran to Los Angeles, London to Dubai—the search term "film action doble farsi 2025" has become a trending beacon of excitement. But what exactly does this phrase mean, and why is it gaining traction?

In this comprehensive guide, we break down the concept of "Double Action" (dubbed/dual-audio) films coming in 2025, analyze the most anticipated blockbusters that will receive Persian dubbing, and explore how the landscape of Farsi-dubbed action movies is evolving.

Why 2025? In the real world, the Iranian film industry faces structural delays: sanctions affecting CGI software, difficulty importing camera rigs for car chases, and a censorship board that scrutinizes violence (gunplay is often minimized, replaced by hand-to-hand combat or symbolic martyrdom).

By 2025, three trends will converge:

What would this 2025 film actually look like? Based on the search query’s energy, the plot likely involves:

Conclusion

In this work, we introduce Voyager, the first LLM-powered embodied lifelong learning agent, which leverages GPT-4 to explore the world continuously, develop increasingly sophisticated skills, and make new discoveries consistently without human intervention. Voyager exhibits superior performance in discovering novel items, unlocking the Minecraft tech tree, traversing diverse terrains, and applying its learned skill library to unseen tasks in a newly instantiated world. Voyager serves as a starting point to develop powerful generalist agents without tuning the model parameters.

Media Coverage

"They Plugged GPT-4 Into Minecraft—and Unearthed New Potential for AI. The bot plays the video game by tapping the text generator to pick up new skills, suggesting that the tech behind ChatGPT could automate many workplace tasks." - Will Knight, WIRED

"The Voyager project shows, however, that by pairing GPT-4’s abilities with agent software that stores sequences that work and remembers what does not, developers can achieve stunning results." - John Koetsier, Forbes

"Voyager, the GTP-4 bot that plays Minecraft autonomously and better than anyone else" - Ruetir

"This AI used GPT-4 to become an expert Minecraft player" - Devin Coldewey, TechCrunch

Coverage Index: [Atmarkit] [Career Engine] [Crast.net] [Daily Top Feeds] [Entrepreneur en Espanol] [Finance Jxyuging] [Forbes] [Forbes Argentina] [Gaming Deputy] [Gearrice] [Haberik] [Head Topics] [InfoQ] [ITmedia News] [Mark Tech Post] [Medium] [MSN] [Note] [Noticias de Hoy] [Ruetir] [Stock HK] [Tech Tribune France] [TechCrunch] [TechBeezer] [Toutiao] [US Times Post] [VN Explorer] [WIRED] [Zaker]

Team

Guanzhi Wang

Yuqi Xie

Yunfan Jiang^*

Ajay Mandlekar^*

Chaowei Xiao

Yuke Zhu

Linxi "Jim" Fan^†

Anima Anandkumar^†

* Equal Contribution † Equal Advising

BibTeX

@article{wang2023voyager,
  title   = {Voyager: An Open-Ended Embodied Agent with Large Language Models},
  author  = {Guanzhi Wang and Yuqi Xie and Yunfan Jiang and Ajay Mandlekar and Chaowei Xiao and Yuke Zhu and Linxi Fan and Anima Anandkumar},
  year    = {2023},
  journal = {arXiv preprint arXiv: Arxiv-2305.16291}
}