All benchmarks

KernelGen 1P

Paper

Progress Over Time

Interactive timeline showing model performance evolution on KernelGen 1P

State-of-the-art frontier
Open
Proprietary

KernelGen 1P Leaderboard

3 models
ContextCostLicense
1
OpenAI
GPT-5.6 SolNew
OpenAI
—1.1M$5.00 / $30.00
2
OpenAI
GPT-5.6 TerraNew
OpenAI
—1.1M$2.50 / $15.00
3
OpenAI
GPT-5.6 LunaNew
OpenAI
—1.1M$1.00 / $6.00
Notice missing or incorrect data?
About this benchmark

What is KernelGen 1P?

KernelGen 1P is an OpenAI AI-self-improvement evaluation that measures whether models can write and optimize compute kernels, part of the suite tracking progress toward accelerating internal research.

KernelGen 1P is a text benchmark evaluating models on agents, code, and systems tasks. LLM Stats tracks 3 models on this benchmark, scored on a 0–1 scale. The current average is 0.4, with the leader at 0.6.

Compare leaders on the best AI for agents, best AI for code and best AI for systems leaderboards.

Current leaders

GPT-5.6 Sol from OpenAI currently leads the KernelGen 1P leaderboard with a score of 0.611 across 3 evaluated AI models.

1GPT-5.6 SolOpenAI61.1%
2GPT-5.6 TerraOpenAI49.2%
3GPT-5.6 LunaOpenAI22.4%

FAQ

Common questions about the KernelGen 1P benchmark and leaderboard.

What is the KernelGen 1P benchmark?

KernelGen 1P is an OpenAI AI-self-improvement evaluation that measures whether models can write and optimize compute kernels, part of the suite tracking progress toward accelerating internal research.

What is the KernelGen 1P leaderboard?

The KernelGen 1P leaderboard ranks 3 AI models based on their performance on this benchmark. Currently, GPT-5.6 Sol by OpenAI leads with a score of 0.611. The average score across all models is 0.442.

What is the highest KernelGen 1P score?

The highest KernelGen 1P score is 0.611, achieved by GPT-5.6 Sol from OpenAI.

How many models are evaluated on KernelGen 1P?

3 models have been evaluated on the KernelGen 1P benchmark, with 0 verified results and 3 self-reported results.

Where can I find the KernelGen 1P paper?

The KernelGen 1P paper is available at https://openai.com/index/gpt-5-6/. The paper details the methodology, dataset construction, and evaluation criteria.

What categories does KernelGen 1P cover?

KernelGen 1P is categorized under agents, code, and systems. The benchmark evaluates text models.

Which model offers the best value on KernelGen 1P?

Among models scoring within 10% of the leader, GPT-5.6 Sol from OpenAI is the cheapest, at $5.00 per million input tokens with a score of 0.611.

How recent are the KernelGen 1P leaderboard results?

The KernelGen 1P leaderboard was last updated in July 2026 and currently includes 3 evaluated models.