Free tools Windows power users keep installed
One-click scans. No signup required.
On August 11, 2017, OpenAI’s self-trained Dota bot defeated professional player Danil “Dendi” Ishutin 2–0 in a live best-of-three, one-versus-one exhibition at The International 2017. Dendi conceded during the second game after falling decisively behind. The “he quits” headline referred to that game or match—not to Dendi retiring from professional Dota 2 or abandoning the game.
What happened at The International 2017?
OpenAI staged the exhibition during Dota 2’s premier tournament. It was a 1-vs.-1 contest played under the tournament rules for that format, not a conventional five-player team match. OpenAI’s account records a 2–0 best-of-three victory; contemporary GamesBeat coverage reported that Dendi threw in the towel during the second game when the bot had built a commanding advantage.
OpenAI’s event account and the contemporary GamesBeat report support the result. The original GamesBeat story was published on August 12, 2017, describing the match from the previous day.
Who was Dendi?
Danil “Dendi” Ishutin was a Ukrainian professional, former world champion and one of Dota 2’s best-known personalities. OpenAI later described him as a 7.3k-rated professional and major fan favorite. That establishes elite status and historical significance, but it does not prove he was unambiguously the world’s number-one player at that precise moment. OpenAI also said the bot had beaten other prominent professionals in the days around the exhibition.
#1 Best Overall
- Authentic Design: Officially inspired Dota 2 arcade game metal sign featuring classic retro marquee styling perfect for gaming enthusiasts
- Premium Materials: Constructed with high-quality aluminum and professional-grade printing for long-lasting durability and vibrant appearance
- Ideal Size: Compact 4" x 11" dimensions make it perfect for mounting on walls or displaying on shelves in your gaming setup
- American Made: Proudly manufactured in the USA with attention to detail and superior craftsmanship
- Versatile Decor: Perfect addition to enhance gaming rooms, PC setups, streaming spaces, or any area where you want to showcase your passion for Dota 2
Accordingly, “top Dota 2 player” is fair shorthand for an elite former champion, while “the best player in the world” would be stronger than the cited evidence supports.
What OpenAI actually built
A specialist for one-versus-one play
The 2017 opponent was an early OpenAI Dota system designed for the game’s 1-vs.-1 setting. It was not OpenAI Five, the later team of five neural networks that played full Dota 2. The distinction matters: a duel removes teammates, coordinated support, drafting and much of the map-wide objective management that define ordinary Dota.
OpenAI’s announcement says the bot learned through self-play reinforcement learning. It repeatedly played against copies of itself, received feedback from game outcomes and adjusted its policy over time. For this system, OpenAI said it did not use imitation learning from human demonstrations or conventional tree search. See OpenAI’s description of the bot.
Its information and actions
According to OpenAI’s technical follow-up, the bot used Valve’s Dota bot API. Its observations covered the hero, creeps, courier and nearby terrain; its actions included moving, attacking a unit and using an item. That is a restricted game interface rather than an unrestricted, omniscient view.
The Tool Desk
Outbyte PC Repair FREEClear out junk files and repair common Windows errorsFree Scan →Outbyte Driver Updater FREEScan for outdated or missing drivers - takes under a minuteDriver Scan →Rank #2
- Size:300*260*3MM Material: Cloth + natural Rubber
- Adapted Natural Pigment coating, improve mouse reaction. Design for pro gaming
- Waterproof, nonskid, Catcher
- One year warranty
- Peru Doto best Doto!!!
“Human-accessible” did not make the conditions identical in every practical respect. A machine can execute decisions with exceptional timing and consistency, and it can train through vastly more games than a person. OpenAI’s detailed account is at More on Dota 2.
Why a 1-vs.-1 victory was still a serious AI result
Dota 2 is difficult for learning systems because it combines several problems at once:
- Real-time decisions rather than alternating turns
- Partial information about the opponent and map
- Large, continuous state and action spaces
- Delayed consequences for resource and positioning choices
- Opponent adaptation and strategic deception
Those properties make Dota a demanding environment even when the rules are narrowed. OpenAI contrasted its real-time, partially observable, multi-agent challenges with games such as chess, Go and Atari in its project overview. The Dendi match therefore demonstrated superhuman performance on a defined subproblem—not general intelligence and not mastery of unrestricted five-on-five Dota.
How dominant was the bot?
The following scores come from OpenAI’s own August 2017 retrospective and should be read as first-party reported results:
What’s actually slowing this PC down?
Pick the symptom - the matching free tool is one click away.
Rank #3
| Opponent | Reported result |
|---|---|
| Blitz, former professional | 3–0 |
| Pajkatt, professional | 2–1 |
| CC&C, professional | 3–0 |
| Arteezy | 10–0 |
| SumaiL | 6–0 |
| Dendi | 2–0 |
OpenAI published these figures, along with Dendi’s historical rating and the bot’s development timeline, in its follow-up report. They show that the Dendi exhibition was part of a broader run against elite players, not a single isolated upset. They are not presented here as independently audited records.
Was the bot unbeatable?
No. OpenAI later documented ways humans could exploit the 1-vs.-1 system, including creep pulling, opening with Orb of Venom and Wind Lace, and forcing a difficult level-one Raze sequence. These tactics exposed a familiar limitation of narrow self-play: a system can dominate the situations represented in training while remaining vulnerable to unusual strategies outside that distribution.
That qualification does not erase the achievement. It explains what the result measured: a highly capable policy in a controlled task, rather than an opponent immune to every novel plan.
How 1-vs.-1 differed from full Dota 2
Standard Dota 2 is a 5-vs.-5 multiplayer game. Team communication, coordinated attacks, support roles, drafting, vision, map control, Roshan and long-term objective choices are central to it. The 2017 duel removed four teammates on each side and substantially narrowed the strategic problem.
Windows Errors? Fix Them Before They Spread
Repair common Windows errors and clear accumulated junk for a smoother, more stable PC - no reinstall needed.Free scan · no reinstallCrashes, No Sound, or Screen Glitches?
Random freezes, missing sound and display glitches usually trace back to one bad driver. Find and replace yours safely.Free scan · under a minuteRank #4
- You can redeem this code to receive a cosmetic upgrade for many of the DOTA 2 characters!
OpenAI explicitly treated the duel as an earlier stage toward the much harder full-game project. Calling the Dendi opponent “OpenAI Five,” or saying it had already solved Dota 2, is historically inaccurate. The 2017 bot and the later five-agent system should be described as different generations of the project.
Independent reader supportYour contribution helps us test, update, and keep practical guides available for everyone.From the Dendi exhibition to OpenAI Five
Scaling to a five-agent team
OpenAI later developed OpenAI Five, a system of five neural networks playing full Dota 2. In a 2018 description, OpenAI said the system trained through approximately 180 years of simulated games per day using 256 GPUs and 128,000 CPU cores. Those figures describe that historical training setup, not a current OpenAI infrastructure claim. See OpenAI Five.
Mixed results in 2018
In an August 2018 benchmark, OpenAI Five won a best-of-three against a team OpenAI described as being in the 99.95th percentile of Dota players. The human side won the third game after an audience-selected adversarial draft, a caveat documented in OpenAI’s benchmark report.
At The International 2018, OpenAI Five lost two matches against stronger professional opposition. OpenAI characterized portions of those games as close, while also distinguishing the event’s restrictions from “Real Dota.” The results are recorded in its International 2018 report.
The 2019 milestone
In April 2019, OpenAI Five defeated reigning Dota 2 world champions OG in two consecutive games. That was a full-game, five-versus-five milestone and a different achievement from the 2017 Dendi duel. OpenAI’s account is available at OpenAI Five defeats Dota 2 world champions.
What the 2017 match actually proved
The exhibition provided strong evidence that large-scale self-play reinforcement learning could produce highly capable behavior in a real-time, partially observable environment. It did not establish general intelligence, eliminate the importance of human-designed interfaces or prove that an AI had mastered every strategic and social dimension of multiplayer games.
The most accurate reading is narrower and more useful: OpenAI built a specialized 1-vs.-1 Dota bot, trained it through self-play, and demonstrated that it could beat elite professionals in public competition. Dendi’s concession made for a dramatic headline, but it was a concession of the match—not a decision to quit Dota 2.
Quick Recap
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
Quick wins for a faster PC:
Scan for outdated or missing drivers - takes under a minuteDriver Scan →Clear out junk files and repair common Windows errorsFree Scan →Fix the driver behind crashes, sound loss and screen glitchesFind Drivers →
