LLM Benchmarks

The Largest Coding Benchmark for Large Language Models [LLM]

---

information of project

all info about project, which llms used in benchmarks, prompts and etc is in docs/

For Evaluators

Additional Guidance

• Request brevity: Add “Solution must be ≤ 300 lines” to each prompt
• Accept pseudocode/core proofs for very long tasks
• Encourage inclusion of minimal unit tests within responses

license

Distributed under the MIT License. See LICENSE file for details.