USACO
The USACO benchmark evaluates AI agents on competitive programming problems from the USA Computing Olympiad. It consists of 307 problems, complete with exhaustive test cases, problem analyses, and difficulty labels.
Paper: Can Language Models Solve Olympiad Programming? (Shi et al., 2024)
Key Features of USACO
Difficulty Levels
Tasks span Bronze to Platinum difficulty levels, requiring knowledge of data structures and algorithms.
Real Competition Tasks
Problems are sourced from actual USACO competitions, consisting of problems challenging for human programmers.
USACO Leaderboard
| Rank | Agent |
Primary Model
Primary Model
This is the primary model used by the agent. In some cases, an embedding model is used for RAG, or a secondary model like GPT-4o for image processing. Note: For non-OpenAI reasoning models, the reasoning token budget is set at 1,024 (low), 2,048 (medium), and 4,096 (high). |
Verified
Verified Results
Results have been reproduced by the HAL team |
Accuracy
Accuracy
Confidence intervals show the min-max values across runs for those agents where multiple runs are available |
Cost (USD)
Total Cost
Total API cost for running the agent on all tasks. Confidence intervals show the min-max values across runs for those agents where multiple runs are available |
Runs
Number of Runs
The number of runs for this agent submitted to the leaderboard. To submit multiple evaluations, rerun the same agent and set the same agent name |
Traces |
|---|---|---|---|---|---|---|---|
| 1 |
USACO Episodic + Semantic
Pareto optimal
|
GPT-5 Medium (August 2025) | ✓ | 69.71% | $64.13 | 1 | Download |
| 2 | o4-mini High (April 2025) | ✓ | 57.98% | $44.04 | 1 | Download | |
| 3 | Claude Opus 4.1 High (August 2025) | ✓ | 51.47% | $267.72 | 1 | Download | |
| 4 | Claude Opus 4.1 (August 2025) | ✓ | 48.21% | $276.19 | 1 | Download | |
| 5 | o3 Medium (April 2025) | ✓ | 46.25% | $57.30 | 1 | Download | |
| 6 | GPT-4.1 (April 2025) | ✓ | 44.95% | $28.10 | 1 | Download | |
| 7 |
USACO Episodic + Semantic
Pareto optimal
|
DeepSeek V3 | ✓ | 39.09% | $2.78 | 1 | Download |
| 8 | DeepSeek R1 | ✓ | 38.11% | $8.18 | 1 | Download | |
| 9 | o4-mini Low (April 2025) | ✓ | 30.94% | $21.14 | 1 | Download | |
| 10 | Claude-3.7 Sonnet (February 2025) | ✓ | 29.32% | $38.70 | 1 | Download | |
| 11 |
USACO Episodic + Semantic
Pareto optimal
|
Gemini 2.0 Flash | ✓ | 27.04% | $1.46 | 1 | Download |
| 12 | Claude-3.7 Sonnet High (February 2025) | ✓ | 26.71% | $56.43 | 1 | Download | |
| 13 | GPT-4.1 (April 2025) | ✓ | 25.41% | $197.33 | 1 | Download |
Accuracy vs. Cost Frontier for USACO
This plot shows the relationship between an agent's performance and its token cost. The Pareto frontier (dashed line) represents the current state-of-the-art trade-off. The error bars indicate min-max values across runs.
Heatmap for USACO
The heatmap visualizes success rates across tasks and agents. Colorscale shows the fraction of times a task was solved across reruns of the same agent. The "any agent" performance indicates the level of saturation of the benchmark and gives a sense of overall progress.
Total Completion Tokens Used per Agent
The bar chart shows the total completion tokens used by each agent, with the height of each bar representing the total number of completion tokens used across all tasks. Secondary models usually contribute a relatively small amount of tokens in comparison, and are used for RAG or image processing only.
Model Performance Over Time
Track how model accuracy has evolved over time since their release dates. Each point represents the best performance achieved by that model on USACO.
Token Pricing Configuration
Adjust token prices to see how they affect the total cost calculations in the leaderboard and plots.
Additional Resources
Getting Started
Want to evaluate your agent on USACO? Follow our comprehensive guide to get started:
View DocumentationTask Details
Browse the complete list of USACO tasks, including difficulty levels and categories:
View Tasks