Contest Archive - AtCoder

AtCoder Beginner Contest 480

Sat, 17 Oct 2026 12:00:00 GMT

Contest Information

Duration: 100 minutes
Rated Range: 0 - 1999

About the New Judge

This contest will be conducted using a new judge. Please thoroughly check the available languages and versions before participating.
For information on available languages, versions, and libraries, please check here.
Display format for TLE and MLE has been changed.
If there are critical issues like InternalError, please notify us in Clar with the relevant submission attached.
We have already closed requests regarding versions and libraries, hence we cannot address them.

Point Values

Task	Score
A	100
B	200
C	300
D	400
E	450
F	525
G	600

Contest Rules

This contest is full-feedback (solutions are judged during the contest).
When you solve a problem, you get a score assigned to it. Competitors are ranked first by total scores, then by penalties. The penalties are computed as (the time you spend to get your current score) + (5 minutes) * (the number of incorrect attempts).

Useful Links

AtCoder Heuristic Contest 073

Sat, 17 Oct 2026 06:00:00 GMT

Contest Information

In short-duration contests, the use of generative AI is now prohibited, with limited exceptions. For details, please see this article.
Writer: wata_admin
You can earn GP30 scores for AWTF2027.
We are currently holding a trial contest to evaluate proposed new generative AI usage rules for future long-term AHCs. We encourage you to participate and share your feedback on the proposed rules.

Rules

There is one problem. You can use any programming language available on AtCoder.
There is no penalty for resubmission, but you must wait at least 5 minutes between submissions.
This is an individual competition. It is prohibited to publish your solution before the end of the contest. However, you are free to share information on how to run the provided tools.
For other rules, see here.

AHC GenAI Rules Trial Contest 2026

Thu, 15 Oct 2026 10:00:00 GMT

Contest Information

This is a trial contest to evaluate proposed new generative AI usage rules for long-duration AHCs. It is unrated, and no GP30 points will be awarded.
The generative AI rules on this page apply to this contest. They differ from the regular rules for short-duration AHCs.
Our goal is to evaluate clarity, practicality, and competitive impact, and to collect feedback.
Please use the hashtag #AHC_GenAI26 when posting about the contest. As with regular contests, sharing solutions during the contest is prohibited. However, you are welcome to share your thoughts and feedback on the generative AI usage rules even while the contest is ongoing.
The submission interval for this contest is 10 minutes.

Background of the Proposal

The generative AI rules previously used in long-duration AHCs primarily restricted large-scale generation using APIs and autonomous improvement loops. However, even under these rules, modern generative AI can arrive at exceptionally strong solutions simply by repeatedly receiving vague instructions such as “improve it further.”

We believe long-duration AHCs currently face two major challenges:

If generative AI is freely allowed, the performance difference attributable to the AI system used may exceed the difference attributable to participants' own ideas for a large portion of the field.
As generative AI reduces opportunities to write code manually in everyday life, banning it entirely could substantially raise the barrier to participation and lead to a large decline in participants.

We have therefore considered splitting long-duration AHCs into a human only division, which bans generative AI, and a with AI division, which accepts pay-to-win disparities and fully permits generative AI, including autonomous improvement loops.

However, generative AI capabilities are improving faster than expected. In chess, there was a time when human–AI teams outperformed AI playing alone, but as AI has advanced, human intervention in choosing moves no longer necessarily improves performance. Similarly, we believe that in the near future, many AHC participants may find that developing and intervening with their own algorithmic ideas no longer improves their results. The contest would then shift from competing on human-designed solutions to competing on systems that use generative AI to solve AHC problems.

For example, in October 2026, OpenAI published a large collection of mathematics research outputs from an unreleased internal model. Individual results may still require further verification, but the announcement illustrates rapid progress in AI's ability to perform sophisticated reasoning.

Developing autonomous AI systems is valuable in its own right. However, this is fundamentally different from AHC's goal of competition between human participants. If the goal is to compare the performance of systems without human intervention, we believe a multi-problem benchmark such as ALE-Bench is a better fit than a single contest.

Even if generative AI eventually achieves higher scores than participants could by developing solutions themselves, AHC aims to remain a place where people can enjoy the process of facing problems, learning, thinking, experimenting, and applying what they discover to their next ideas. We want to preserve that enjoyment, even in an era when generative AI far surpasses human capabilities.

Ideally, the organizers would provide an official AI under identical conditions for every participant. At present, its cost makes that impractical. As a compromise, we are considering a rule under which humans remain responsible for developing and improving solutions, while generative AI may assist with implementation.

Even under these rules, it is impossible to define a perfectly precise boundary between permitted and prohibited uses. As a practical compromise, we accept that differences in how this boundary is interpreted may have some effect on scores and standings. However, given the complexity of long-duration AHC problems, we expect such differences in borderline AI use to rarely have a major impact on rankings.

This contest is a trial intended to evaluate these proposed rules and gather participant feedback. It does not mean that the rules have already been adopted for future regular AHCs.

Generative AI Usage Rules (Proposed; Applied for This Trial)

1. Basic Principle

These rules apply during AtCoder Heuristic Contests (AHCs) that explicitly adopt them.

Participants must make their own decisions about developing and improving solutions (referred to below as “solution-level decisions”). Generative AI may be used to assist with implementation.

“Solution-level decisions” include devising, comparing, selecting, or changing algorithms, search strategies, neighborhoods, objective or evaluation functions, search parameters, experiment plans, and improvement strategies derived from experimental results. The mechanical procedures expressly permitted in Section 2 are excluded.

2. Permitted Uses

The following uses are permitted:

Implementing a solution specified by the participant or already defined in existing code.
Investigating and fixing discrepancies between the specified solution and its implementation, as well as implementation errors.
Optimizing performance or refactoring code without changing the solution.
Repeating testing, measurement, and fixes necessary for the above implementation work.
Running experiments specified by the participant; mechanically aggregating, statistically analyzing, and visualizing results; and mechanically comparing, searching, or selecting parameters or solution candidates using candidate values, ranges, and evaluation criteria specified by the participant. However, AI must not devise new solution candidates or improvement strategies.
Creating supporting tools such as visualizers.
Explaining the current source code.
Receiving explanations of general algorithmic knowledge and how algorithms work. However, asking generative AI to analyze how an algorithm should be applied to the contest problem or how effective it would be for that problem is prohibited.
Translating problem statements, organizing input/output specifications, and obtaining answers to programming questions.

The scope of permitted implementation work is the same whether generative AI implements a participant-specified solution from scratch or modifies existing code afterward. Accordingly, broad instructions such as “Check whether the code implements the specified solution correctly, and fix any bugs you find” or “optimize the implementation without changing the specified solution or algorithm” are permitted without specifying the exact parts to change or the methods to use, regardless of whether the code was written by the participant or generated by AI. However, the AI must not make solution-level decisions during these tasks. If any solution-level details necessary for the work are unclear, the participant must decide them rather than having the AI determine them independently. For example, asking AI to “analyze why this solution scores poorly” without limiting the investigation to implementation errors, and thereby having it examine shortcomings of the solution itself, is prohibited.

3. Prohibited Uses

Participants must not have generative AI make solution-level decisions. Prohibited uses include:

Having AI devise, suggest, compare, or select solutions or improvement ideas, except for mechanical comparison or selection expressly permitted under Section 2.
Having AI decide solution-level details that have not been specified.
Having AI analyze experimental results or scores to identify shortcomings of the solution itself or decide how to improve it.

Both directly requesting such conduct and intentionally inducing the AI to perform it are prohibited. For example, if the AI asks for missing solution-level details, instructing it to “figure it out yourself” or “pick the best approach” is prohibited.

A participant's instruction being incomplete or ambiguous does not, by itself, constitute a violation.

4. Mandatory Instructions to Generative AI

Participants must apply the organizer-specified mandatory instructions (which include instructions for the AI and the full text of these rules) to every generative AI used in connection with the contest. This may be done through a message at the start of a chat, custom instructions, AGENTS.md, CLAUDE.md, or another method that the AI can access. If multiple AI services or independent sessions are used, the instructions must be applied to each.

Participants must not intentionally disable or circumvent the mandatory instructions. The exceptions in Section 6 do not require these instructions.

5. Unintentional Violations by AI

If generative AI performs a prohibited action without being instructed or induced to do so by the participant, that fact alone does not constitute a violation by the participant.

If the AI detects a prohibited action before presenting the resulting work to the participant and can safely undo it, the instructions require it to undo the prohibited changes without presenting that work. On the other hand, the participant may use ideas, code, and other outputs that have already been presented or cannot safely be undone. The participant need not discard such outputs or revert code.

If the participant becomes aware of prohibited behavior, they must provide the mandatory instructions specified in Section 4 again, tell the AI that its behavior was prohibited, and instruct it not to repeat the behavior.

If the participant recognizes a high likelihood that the AI will repeatedly perform prohibited actions, they must take reasonable steps to prevent recurrence, such as revising their instructions or usage practices, switching to a different model, or, where necessary, discontinuing that use.

If the participant has taken appropriate measures to prevent recurrence but the AI nevertheless performs a prohibited action, that fact alone does not constitute a rule violation by the participant.

This section protects participants from unintended AI behavior; it does not authorize the AI to perform prohibited actions.

6. Special Cases

Code completion: Generative-AI code-completion features expressly permitted by the Short-Term AtCoder Heuristic Contest Generative AI Usage Rules — Version 20260818 are also permitted here. Mandatory AI instructions and log preservation/submission are not required for these specified features. Other features of the same products are not automatically exempt.

AI-generated overviews in web search: These may be viewed and used as part of ordinary web searches. Mandatory AI instructions and log preservation/submission are not required for this use. However, using them to circumvent these rules is prohibited. For example, searching for “simulated annealing” and reading an AI overview is permitted, but pasting the contest problem into a search box and asking it to “come up with a solution to the following problem” in order to obtain an AI-generated solution is prohibited. Interactive AI-search features are treated as ordinary generative-AI use and must comply with these rules.

Uses unrelated to the contest: Generative AI may be used for work, education, personal matters, or other purposes unrelated to the contest. These uses are not subject to the restrictions or logging requirements here.

7. Preservation and Submission of AI Usage Logs

Participants must preserve conversations and available work logs for generative AI used in connection with the contest, except for the uses exempted by Section 6.

Participants eligible to earn GP30 points, receive prize money, or qualify for a subsequent round must submit their AI usage logs. Failure to submit the required logs by the stated deadline results in loss of the corresponding eligibility.

Only logs related to the contest need to be submitted. Logs of AI use for unrelated work, education, or personal matters are not required.

Personal information and confidential information irrelevant to the review may be redacted. However, instructions to and responses from AI, and any other information needed to assess rule compliance, must not be concealed. Intentional falsification or concealment of review-relevant information is prohibited.

Usage logs include the prompts sent to AI, AI responses, and available tool-execution records. Internal reasoning or other information not made available to the user need not be submitted.

How to retrieve logs depends on the tool. For example, when using Codex CLI, participants can ask Codex, “I need to submit logs under the contest rules. Find the Codex session logs started from this directory,” to help locate the saved records. Participants must check that the collected logs cover all relevant usage.

Participants who did not use AI, or used only features exempt from logging, may declare that fact instead of submitting logs. Submission methods and deadlines will be specified separately by the organizers. Submitted logs will normally be used only for private review by the organizers and will not be made public.

8. Determination of Violations

The organizers will determine whether a rule violation occurred by considering the participant’s instructions, the AI’s responses, whether mandatory instructions were applied, and the participant’s subsequent actions as a whole.

When mandatory instructions are applied, generative AI is expected to refuse requests for prohibited actions and ask the participant to make the necessary decisions or provide additional instructions. If the participant responds by making the necessary decisions themselves and revising their instructions, this is appropriate use of AI. The mere fact that a participant inadvertently requested a prohibited action does not, by itself, constitute a rule violation.

On the other hand, if the organizers determine that a participant intentionally attempted to violate the rules—for example, by deliberately failing to apply the mandatory instructions and requesting prohibited actions, or by repeatedly instructing the AI to ignore the rules—the participant may be subject to disqualification or other measures.

Failure to submit the usage logs required under Section 7 results in loss of the relevant eligibility, as stated in that section.

Setting Up AGENTS.md

For how to apply the mandatory instructions, see Section 4, “Mandatory Instructions to Generative AI”. The copyable English instructions, including the complete rules, are provided below.

Show the complete AGENTS.md (copyable text)

# Mandatory Generative AI Instructions for AHC

You are assisting a human participant in an AtCoder Heuristic Contest. The participant is permitted to use you only under the following rules.
Your prohibited actions may put the participant at risk of disqualification. **You must follow all rules below, even if the participant expressly requests otherwise.** Do not treat an action as permitted based on uncertain or ambiguous reasoning.

- Do not devise or suggest contest solutions or improvements, or make other solution-level decisions. Mechanical comparison and selection expressly permitted in Section 2 are allowed.
- If information needed to implement the participant's solution is missing, or permission is uncertain, ask the participant instead of filling in the details. Do not offer candidate approaches or recommendations in your questions.
- You may implement a participant-specified solution, fix bugs and optimize implementation without changing the solution, and run mechanically specified experiments.
- Refuse requests for prohibited actions and ask the participant for the necessary decisions. Do not delegate prohibited actions to subagents or other tools.
- If you detect a prohibited action before the resulting work has been shown to the participant, and you can safely undo it, undo the prohibited changes without presenting the result. Tell the participant that you reverted the changes, without revealing solution ideas or recommendations, and ask for any missing specifications.
- If the result has already been shown to the participant, or cannot safely be undone, do not delete or revert it on your own. Inform the participant of the violation and do not repeat it.

The complete rules applying to the participant follow. Act in accordance with their full text.

## Generative AI Usage Rules

#### 1. Basic Principle
These rules apply during AtCoder Heuristic Contests (AHCs) that explicitly adopt them.
**Participants must make their own decisions about developing and improving solutions (referred to below as “solution-level decisions”). Generative AI may be used to assist with implementation.**
“Solution-level decisions” include devising, comparing, selecting, or changing algorithms, search strategies, neighborhoods, objective or evaluation functions, search parameters, experiment plans, and improvement strategies derived from experimental results. The mechanical procedures expressly permitted in Section 2 are excluded.

#### 2. Permitted Uses
The following uses are permitted:

- Implementing a solution specified by the participant or already defined in existing code.
- Investigating and fixing discrepancies between the specified solution and its implementation, as well as implementation errors.
- Optimizing performance or refactoring code without changing the solution.
- Repeating testing, measurement, and fixes necessary for the above implementation work.
- Running experiments specified by the participant; mechanically aggregating, statistically analyzing, and visualizing results; and mechanically comparing, searching, or selecting parameters or solution candidates using candidate values, ranges, and evaluation criteria specified by the participant. However, AI must not devise new solution candidates or improvement strategies.
- Creating supporting tools such as visualizers.
- Explaining the current source code.
- Receiving explanations of general algorithmic knowledge and how algorithms work. However, asking generative AI to analyze how an algorithm should be applied to the contest problem or how effective it would be for that problem is prohibited.
- Translating problem statements, organizing input/output specifications, and obtaining answers to programming questions.

The scope of permitted implementation work is the same whether generative AI implements a participant-specified solution from scratch or modifies existing code afterward. Accordingly, broad instructions such as “Check whether the code implements the specified solution correctly, and fix any bugs you find” or “optimize the implementation without changing the specified solution or algorithm” are permitted without specifying the exact parts to change or the methods to use, regardless of whether the code was written by the participant or generated by AI. However, the AI must not make solution-level decisions during these tasks. If any solution-level details necessary for the work are unclear, the participant must decide them rather than having the AI determine them independently. For example, asking AI to “analyze why this solution scores poorly” without limiting the investigation to implementation errors, and thereby having it examine shortcomings of the solution itself, is prohibited.

#### 3. Prohibited Uses
**Participants must not have generative AI make solution-level decisions.** Prohibited uses include:

- Having AI devise, suggest, compare, or select solutions or improvement ideas, except for mechanical comparison or selection expressly permitted under Section 2.
- Having AI decide solution-level details that have not been specified.
- Having AI analyze experimental results or scores to identify shortcomings of the solution itself or decide how to improve it.

Both directly requesting such conduct and intentionally inducing the AI to perform it are prohibited. For example, if the AI asks for missing solution-level details, instructing it to “figure it out yourself” or “pick the best approach” is prohibited.
A participant's instruction being incomplete or ambiguous does not, by itself, constitute a violation.

#### 4. Mandatory Instructions to Generative AI
Participants must apply the organizer-specified mandatory instructions (which include instructions for the AI and the full text of these rules) to every generative AI used in connection with the contest. This may be done through a message at the start of a chat, custom instructions, `AGENTS.md`, `CLAUDE.md`, or another method that the AI can access. If multiple AI services or independent sessions are used, the instructions must be applied to each.
Participants must not intentionally disable or circumvent the mandatory instructions. The exceptions in Section 6 do not require these instructions.

#### 5. Unintentional Violations by AI
If generative AI performs a prohibited action without being instructed or induced to do so by the participant, that fact alone does not constitute a violation by the participant.
If the AI detects a prohibited action before presenting the resulting work to the participant and can safely undo it, the instructions require it to undo the prohibited changes without presenting that work. On the other hand, **the participant may use ideas, code, and other outputs that have already been presented or cannot safely be undone.** The participant need not discard such outputs or revert code.
If the participant becomes aware of prohibited behavior, they must **provide the mandatory instructions specified in Section 4 again**, tell the AI that its behavior was prohibited, and instruct it not to repeat the behavior.
If the participant recognizes a high likelihood that the AI will repeatedly perform prohibited actions, they must take reasonable steps to prevent recurrence, such as revising their instructions or usage practices, switching to a different model, or, where necessary, discontinuing that use.
If the participant has taken appropriate measures to prevent recurrence but the AI nevertheless performs a prohibited action, that fact alone does not constitute a rule violation by the participant.
This section protects participants from unintended AI behavior; it does not authorize the AI to perform prohibited actions.

#### 6. Special Cases
**Code completion:** Generative-AI code-completion features expressly permitted by the [Short-Term AtCoder Heuristic Contest Generative AI Usage Rules — Version 20260818](https://info.atcoder.jp/entry/short-ahc-llm-rules-en) are also permitted here. Mandatory AI instructions and log preservation/submission are not required for these specified features. **Other features of the same products are not automatically exempt.**
**AI-generated overviews in web search:** These may be viewed and used as part of ordinary web searches. Mandatory AI instructions and log preservation/submission are not required for this use. However, using them to circumvent these rules is prohibited. For example, searching for “simulated annealing” and reading an AI overview is permitted, but pasting the contest problem into a search box and asking it to “come up with a solution to the following problem” in order to obtain an AI-generated solution is prohibited. Interactive AI-search features are treated as ordinary generative-AI use and must comply with these rules.
**Uses unrelated to the contest:** Generative AI may be used for work, education, personal matters, or other purposes unrelated to the contest. These uses are not subject to the restrictions or logging requirements here.

#### 7. Preservation and Submission of AI Usage Logs
Participants must preserve conversations and available work logs for generative AI used in connection with the contest, except for the uses exempted by Section 6.
**Participants eligible to earn GP30 points, receive prize money, or qualify for a subsequent round must submit their AI usage logs. Failure to submit the required logs by the stated deadline results in loss of the corresponding eligibility.**
Only logs related to the contest need to be submitted. Logs of AI use for unrelated work, education, or personal matters are not required.
Personal information and confidential information irrelevant to the review may be redacted. However, instructions to and responses from AI, and any other information needed to assess rule compliance, must not be concealed. Intentional falsification or concealment of review-relevant information is prohibited.
Usage logs include the prompts sent to AI, AI responses, and available tool-execution records. Internal reasoning or other information not made available to the user need not be submitted.
How to retrieve logs depends on the tool. For example, when using Codex CLI, participants can ask Codex, “I need to submit logs under the contest rules. Find the Codex session logs started from this directory,” to help locate the saved records. Participants must check that the collected logs cover all relevant usage.
Participants who did not use AI, or used only features exempt from logging, may declare that fact instead of submitting logs. Submission methods and deadlines will be specified separately by the organizers. Submitted logs will normally be used only for private review by the organizers and will not be made public.

#### 8. Determination of Violations
The organizers will determine whether a rule violation occurred by considering the participant’s instructions, the AI’s responses, whether mandatory instructions were applied, and the participant’s subsequent actions as a whole.
When mandatory instructions are applied, generative AI is expected to refuse requests for prohibited actions and ask the participant to make the necessary decisions or provide additional instructions. If the participant responds by making the necessary decisions themselves and revising their instructions, this is appropriate use of AI. **The mere fact that a participant inadvertently requested a prohibited action does not, by itself, constitute a rule violation.**
On the other hand, if the organizers determine that a participant intentionally attempted to violate the rules—for example, by deliberately failing to apply the mandatory instructions and requesting prohibited actions, or by repeatedly instructing the AI to ignore the rules—the participant may be subject to disqualification or other measures.
Failure to submit the usage logs required under Section 7 results in loss of the relevant eligibility, as stated in that section.

Request for Feedback

We are collecting feedback through the “Registration Information” section of this contest. We appreciate your cooperation.

AAL Contest 001: Let's use segtree!

Mon, 12 Oct 2026 04:00:00 GMT

Contest Information

AAL Contest 001: Let's use segtree! is an official educational contest focused on segment trees.
The contest will be held for five hours, from 13:00 to 18:00 on October 12, 2026 (a public holiday), and will consist of 24 problems.
Problem statements will be provided in both Japanese and English.
This contest is unrated. There will be no rating updates.
writers: Nyaan , maspy
tester: Nachia

What Is the AAL Contest?

In addition to courses in the form of explanatory articles and videos, AtCoder Algorithm Lectures will host AAL Contests as educational resources that allow participants to study specific topics in greater depth and gain practical experience.
We hope that you will use these contests to test your skills on the day of the contest and as learning materials afterward, together with the Algorithm Lectures courses.

This Contest's Topic

The topic of the first contest is segment trees.
Segment trees are known as one of the most frequently used data structures among those that are not built into programming languages. They can process aggregate queries in logarithmic time, and by carefully choosing the monoid they store, they can maintain many different kinds of information.
This contest features problems involving many different kinds of information stored in segment trees. Those who are not yet familiar with segment trees are encouraged to work through the problems and learn the techniques for designing the information maintained by a segment tree.
In addition, every problem in this contest can be solved using the segment tree or lazy segment tree provided by AtCoder Library. In other words, extended segment-tree data structures, such as persistent segment trees and Segment Tree Beats!, are outside the scope of this contest. Those who do not have their own segment tree library are encouraged to use the official library.
The problems from Problem R onward require not only knowledge but also substantial problem-solving insight. Advanced participants are also encouraged to take on the contest!

Scoring

Problems A through X are worth a total of 100 points.

Problem	A	B	C	D	E	F	G	H	I	J	K	L	M	N	O	P	Q	R	S	T	U	V	W	X
Score	3	3	3	3	3	3	3	4	4	4	4	4	4	4	4	4	5	5	5	5	5	6	6	6

As a general guideline, participants may aim for approximately the following scores according to their skill level.

Those who are not yet familiar with using segment trees: First, aim to solve the first four problems.
Those who can solve medium-difficulty segment tree problems at around the ABC 500-point level: 25 points
Those who can solve difficult segment tree problems at around the ABC 600-point level: 50 points
Those who are highly proficient with segment trees: 70 points
Those who have completely mastered segment trees: 85 points

Contest Rules

Useful Links

AtCoder Beginner Contest 479

Sun, 11 Oct 2026 12:00:00 GMT

Contest Information

Duration: 100 minutes
Rated Range: 0 - 1999

Promote Your Company through AtCoder Contests

AtCoder is one of the world's largest competitive programming platforms, with over 900,000 registered users worldwide. Our contests attract thousands of highly skilled programmers, students, and engineers.

Sponsorship Benefits

Introduce your company on the contest page
Post a link to your recruitment or corporate website
Reach talented programmers around the world
Receive participant information only from users who agree to provide it to the sponsor

AtCoder offers several contest formats, including AtCoder Beginner Contest (ABC), AtCoder Regular Contest (ARC), and AtCoder Heuristic Contest (AHC).

If you are interested in sponsoring an AtCoder contest, please note that communication will be conducted via text-based communication, such as e-mail or chat.

For more details, please visit: AtCoderContest.pdf

About the New Judge

This contest will be conducted using a new judge. Please thoroughly check the available languages and versions before participating.
For information on available languages, versions, and libraries, please check here.
Display format for TLE and MLE has been changed.
If there are critical issues like InternalError, please notify us in Clar with the relevant submission attached.
We have already closed requests regarding versions and libraries, hence we cannot address them.

Point Values

Task	Score
A	100
B	200
C	300
D	400
E	450
F	500
G	600

Contest Rules

Useful Links

AtCoder Regular Contest++ 232

Sat, 10 Oct 2026 12:00:00 GMT

Contest Information

Duration: 150 minutes
Writer: maroonrk
Rated Range: 1600 - 2999

Point Values

Task	Score
A	800
B	800
C	800
D	800
E	1100

Contest Rules

Useful Links

AtCoder Regular Contest 231

Sun, 04 Oct 2026 12:00:00 GMT

Contest Information

Duration: 120 minutes
Writer: Mitsubachi
Rated Range: 1200 - 2799

Point Values

Task	Score
A	500
B	500
C	600
D	700
E	800
F	800

Contest Rules

Useful Links

AtCoder Beginner Contest 478

Sat, 03 Oct 2026 12:00:00 GMT

Contest Information

Duration: 100 minutes
Rated Range: 0 - 1999

Promote Your Company through AtCoder Contests

Sponsorship Benefits

Introduce your company on the contest page
Post a link to your recruitment or corporate website
Reach talented programmers around the world
Receive participant information only from users who agree to provide it to the sponsor

AtCoder offers several contest formats, including AtCoder Beginner Contest (ABC), AtCoder Regular Contest (ARC), and AtCoder Heuristic Contest (AHC).

If you are interested in sponsoring an AtCoder contest, please note that communication will be conducted via text-based communication, such as e-mail or chat.

For more details, please visit: AtCoderContest.pdf

About the New Judge

This contest will be conducted using a new judge. Please thoroughly check the available languages and versions before participating.
For information on available languages, versions, and libraries, please check here.
Display format for TLE and MLE has been changed.
If there are critical issues like InternalError, please notify us in Clar with the relevant submission attached.
We have already closed requests regarding versions and libraries, hence we cannot address them.

Point Values

Task	Score
A	100
B	200
C	300
D	400
E	475
F	525
G	625

Contest Rules

Useful Links

AtCoder Grand Contest 078

Sun, 27 Sep 2026 12:00:00 GMT

Submitting a recording will be required in order to earn a GP30 score. Click here for details.

Contest Information

Duration: 210 minutes
Writer: yosupo
Rated Range: 2800 - inf

Point Values

Task	Score
A	1500
B	1500
C	1500

Contest Rules

Useful Links

UNICORNProgramming Contest2026(AtCoder Beginner Contest 477)

Sat, 26 Sep 2026 12:00:00 GMT

Contest Information

Duration: 100 minutes
Rated Range: 0 - 1999

Promote Your Company through AtCoder Contests

Sponsorship Benefits

Introduce your company on the contest page
Post a link to your recruitment or corporate website
Reach talented programmers around the world
Receive participant information only from users who agree to provide it to the sponsor

AtCoder offers several contest formats, including AtCoder Beginner Contest (ABC), AtCoder Regular Contest (ARC), and AtCoder Heuristic Contest (AHC).

If you are interested in sponsoring an AtCoder contest, please note that communication will be conducted via text-based communication, such as e-mail or chat.

For more details, please visit: AtCoderContest.pdf

About the New Judge

This contest will be conducted using a new judge. Please thoroughly check the available languages and versions before participating.
For information on available languages, versions, and libraries, please check here.
Display format for TLE and MLE has been changed.
If there are critical issues like InternalError, please notify us in Clar with the relevant submission attached.
We have already closed requests regarding versions and libraries, hence we cannot address them.

Point Values

Task	Score
A	100
B	200
C	300
D	400
E	450
F	525
G	625

Contest Rules

Useful Links

ALGO ARTIS Programming Contest 2026 September（AtCoder Heuristic Contest 072）

Fri, 25 Sep 2026 10:10:00 GMT

Contest Information

Writer: wata_admin
You can earn GP30 scores for AWTF2027.
We have delayed the start time by ten minutes. The end time will remain unchanged. We sincerely apologize for the inconvenience.

Rules

There is one problem. You can use any programming language available on AtCoder.
There is no penalty for resubmission, but you must wait at least 30 minutes between submissions.
This is an individual competition. It is prohibited to publish your solution before the end of the contest. However, you are free to share information on how to run the provided tools.
The ranking results displayed during the contest are provisional. After the contest is over, the system tests against more inputs will be run. The inputs used for provisional ranking will not be used for the system tests. The system test will be performed only for the last submission which received a result other than CE.
There are restrictions on certain uses of generative AI. Please refer to the following document for details: AtCoder Heuristic Contest Generative AI Usage Rules - Version 20250616.
For other rules, see here.

Use of Generative AI

Under the current “AtCoder Heuristic Contest Generative AI Usage Rules — Version 20250616,” it is prohibited to have an AI agent execute a solution program and automatically repeat improvements based on the execution results.

However, with the latest generative AI systems, the distinction between conversational services such as ChatGPT and AI agents such as Codex has become increasingly blurred. Even conversational services may, without an explicit instruction from the user, internally generate test cases, execute a solution program, and automatically repeat improvements based on the execution results.

Some participants already explicitly instruct generative AI systems not to engage in such behavior. To ensure fairness among participants and to enforce the intent of the current rules, anyone using generative AI in this contest must provide the instruction below either at the beginning of each chat or through a custom instruction, project instruction, or instruction file automatically loaded by the relevant generative AI tool, such as AGENTS.md or CLAUDE.md.

I am currently participating in an AtCoder Heuristic Contest, and I will use this generative AI to assist in developing my solution.

When using this generative AI, the "AtCoder Heuristic Contest Generative AI Usage Rules - Version 20250616" apply.

https://info.atcoder.jp/entry/ahc-llm-rules-en

Most importantly, after running the solution program, you must not modify or improve the solution, its approach, or its code based on the execution results unless the user gives a new explicit instruction to do so.

You may run the solution program and report its execution results, logs, scores, or other observations. After reporting them, you must stop and wait for a new instruction from the user before making any improvement based on those results.

Here, "solution program" refers to any program created or being created for the purpose of solving this contest problem, regardless of whether it was created by the user or by generative AI, and regardless of whether it is still in progress or already complete.

AtCoder Regular Contest++ 230

Sun, 20 Sep 2026 12:00:00 GMT

Contest Information

Duration: 150 minutes
Writer: noya2
Rated Range: 1600 - 2999

Point Values

Task	Score
A	600
B	800
C	800
D	900
E	1200

Contest Rules

Useful Links

JIJ Programming Contest 2026（AtCoder Beginner Contest 476）

Sat, 19 Sep 2026 12:00:00 GMT

Contest Information

Duration: 100 minutes
Rated Range: 0 - 1999

About the New Judge

This contest will be conducted using a new judge. Please thoroughly check the available languages and versions before participating.
For information on available languages, versions, and libraries, please check here.
Display format for TLE and MLE has been changed.
If there are critical issues like InternalError, please notify us in Clar with the relevant submission attached.
We have already closed requests regarding versions and libraries, hence we cannot address them.

Point Values

Task	Score
A	100
B	200
C	300
D	425
E	450
F	525
G	625

Contest Rules

Useful Links

AtCoder Heuristic Contest 071

Sun, 13 Sep 2026 10:00:00 GMT

Contest Information

In short-duration contests, the use of generative AI is now prohibited, with limited exceptions. For details, please see this article. Starting with this contest, viewing “AI-generated overviews” in search results has also been added to the list of prohibited activities.
Writer: wata_admin
You can earn GP30 scores for AWTF2027.

Rules

There is one problem. You can use any programming language available on AtCoder.
There is no penalty for resubmission, but you must wait at least 5 minutes between submissions.
This is an individual competition. It is prohibited to publish your solution before the end of the contest. However, you are free to share information on how to run the provided tools.
For other rules, see here.

AtCoder Beginner Contest 475

Sat, 12 Sep 2026 12:00:00 GMT

Contest Information

Duration: 100 minutes
Rated Range: 0 - 1999

Promote Your Company through AtCoder Contests

Sponsorship Benefits

Introduce your company on the contest page
Post a link to your recruitment or corporate website
Reach talented programmers around the world
Receive participant information only from users who agree to provide it to the sponsor

AtCoder offers several contest formats, including AtCoder Beginner Contest (ABC), AtCoder Regular Contest (ARC), and AtCoder Heuristic Contest (AHC).

If you are interested in sponsoring an AtCoder contest, please note that communication will be conducted via text-based communication, such as e-mail or chat.

For more details, please visit: AtCoderContest.pdf

About the New Judge

This contest will be conducted using a new judge. Please thoroughly check the available languages and versions before participating.
For information on available languages, versions, and libraries, please check here.
Display format for TLE and MLE has been changed.
If there are critical issues like InternalError, please notify us in Clar with the relevant submission attached.
We have already closed requests regarding versions and libraries, hence we cannot address them.

Point Values

Task	Score
A	100
B	200
C	300
D	400
E	475
F	525
G	575

Contest Rules

Useful Links

第七回日本最強プログラマー学生選手権～Advance～

Sun, 06 Sep 2026 04:10:00 GMT

AtCoder Beginner Contest 474

Sun, 06 Sep 2026 04:10:00 GMT

Contest Information

Duration: 100 minutes
Rated Range: 0 - 1999

Promote Your Company through AtCoder Contests

Sponsorship Benefits

Introduce your company on the contest page
Post a link to your recruitment or corporate website
Reach talented programmers around the world
Receive participant information only from users who agree to provide it to the sponsor

AtCoder offers several contest formats, including AtCoder Beginner Contest (ABC), AtCoder Regular Contest (ARC), and AtCoder Heuristic Contest (AHC).

If you are interested in sponsoring an AtCoder contest, please note that communication will be conducted via text-based communication, such as e-mail or chat.

For more details, please visit: AtCoderContest.pdf

About the New Judge

This contest will be conducted using a new judge. Please thoroughly check the available languages and versions before participating.
For information on available languages, versions, and libraries, please check here.
Display format for TLE and MLE has been changed.
If there are critical issues like InternalError, please notify us in Clar with the relevant submission attached.
We have already closed requests regarding versions and libraries, hence we cannot address them.

Point Values

Task	Score
A	100
B	200
C	300
D	400
E	450
F	500
G	575

Contest Rules

Useful Links

AtCoder Regular Contest-- 229

Sat, 05 Sep 2026 12:00:00 GMT

Contest Information

Duration: 120 minutes
Writer: sounansya, Nyaan, snuke
Rated Range: 800 - 2399

Point Values

Task	Score
A	400
B	500
C	500
D	600
E	600
F	600

Contest Rules

Useful Links

AtCoder Regular Contest++ 228

Sun, 30 Aug 2026 12:00:00 GMT

Contest Information

Duration: 150 minutes
Writer: PCTprobability
Rated Range: 1600 - 2999

Point Values

Task	Score
A	800
B	800
C	900
D	1000
E	1200