![\"Cover\"](\"https://blog.comma.ai/img/chestnut/cover.jpg\")
Upgrade your comma four’s compute to run openpilot’s biggest models.
\n
Today, we’re launching chestnut, a compute upgrade for comma four. Want to run big models? Nothing beats a desktop GPU. We’re also launching the first chestnut-class model, which is bigger than all openpilot driving models we’ve released combined.
\n![\"chestnut](\"https://blog.comma.ai/img/chestnut/ready_34.png\")
![\"left](\"https://blog.comma.ai/img/chestnut/ready_left.png\")
![\"front](\"https://blog.comma.ai/img/chestnut/ready_front.png\")
![\"right](\"https://blog.comma.ai/img/chestnut/ready_right.png\")
![\"back](\"https://blog.comma.ai/img/chestnut/ready_back.png\")
Since the beginning, openpilot has led on intelligence density. For the last 10 years, we’ve shipped increasingly capable driving models in an on-device compute footprint of 10 W. We shipped fully end-to-end lateral planning, experimental end-to-end longitudinal planning, and even a point-to-point demo to Taco Bell.
\nOver the last two years, we rewrote our stack to scale efficiently with compute and data. With the MLSim architecture, more driving data and training compute translate directly into better openpilot models. The comma fleet now logs nearly one million driving minutes every day, and our scaled-up data pipeline ingests 15% of those drives into the training cluster.
\nBigger models need more compute to run. With chestnut, openpilot is scaling beyond cellphone-sized inference compute. A comma four paired with chestnut has similar compute to Tesla HW4.
\nWe’re launching chestnut with a new 1B parameter driving model in the upcoming openpilot 0.11.2 release. It’s our biggest model yet, and the first in a new line of chestnut-class models.
\n![\"Bignuts\"](\"https://blog.comma.ai/img/chestnut/bignuts.png\")
Compared to the latest on-device model, the first chestnut-class model has 30x more parameters, 100x more FLOPs, and room to go even bigger. Going forward, our autonomy team will focus primarily on maximizing the capabilities of chestnut-class models, and we expect this new class of model to increase development velocity.
\nchestnut comes in a convenient ready to drive kit, which includes chestnut, an AMD Radeon RX 9060 8GB, and everything needed to power, connect, and mount it in your car.
\n![\"Comma](\"https://blog.comma.ai/img/chestnut/comma_four_windshield.png\")
The GPU mounts out of the way in the passenger footwell or under the passenger seat. Run the included cable to comma four, connect to your car’s cigarette lighter, and you’re ready to run openpilot’s biggest models.
\nchestnut is available today in the comma shop.
\nExcited about scaling up? Join us in the #hw-chestnut channel on Discord!
\nWant to help scale up openpilot? We’re hiring for autonomy, product, and operations.
\n", "content_markdown": "*Upgrade your comma four’s compute to run openpilot’s biggest models.*\n\n\n\nToday, we’re launching [chestnut](https://comma.ai/shop/chestnut), a compute upgrade for comma four. Want to run big models? Nothing beats a desktop GPU. We’re also launching the first chestnut-class model, which is bigger than all openpilot driving models we’ve released combined.\n\n    \n\n### Scaling beyond intelligence density [Permalink](https://blog.comma.ai/chestnut/#scaling-beyond-intelligence-density \"Permalink\")\n\nSince the beginning, openpilot has led on intelligence density. For the last 10 years, we’ve shipped increasingly capable driving models in an on-device compute footprint of 10 W. We shipped fully [end-to-end lateral planning](https://blog.comma.ai/end-to-end-lateral-planning/), experimental [end-to-end longitudinal planning](https://blog.comma.ai/090release/#experimental-mode), and even a point-to-point demo to [Taco Bell](https://blog.comma.ai/taco-bell).\n\nOver the last two years, we rewrote our stack to scale efficiently with compute and data. With the [MLSim architecture](https://blog.comma.ai/mlsim), more driving data and training compute translate directly into better openpilot models. The comma fleet now logs nearly one million driving minutes every day, and our scaled-up data pipeline ingests 15% of those drives into the training cluster.\n\nBigger models need more compute to run. With chestnut, openpilot is scaling beyond cellphone-sized inference compute. A comma four paired with chestnut has similar compute to Tesla HW4.\n\n#### A new 1B driving model [Permalink](https://blog.comma.ai/chestnut/#a-new-1b-driving-model \"Permalink\")\n\nWe’re launching chestnut with a new 1B parameter driving model in the upcoming openpilot 0.11.2 release. It’s our biggest model yet, and the first in a new line of chestnut-class models.\n\n\n\nCompared to the latest on-device model, the first chestnut-class model has 30x more parameters, 100x more FLOPs, and room to go even bigger. Going forward, our autonomy team will focus primarily on maximizing the capabilities of chestnut-class models, and we expect this new class of model to increase development velocity.\n\n#### Ready to Drive [Permalink](https://blog.comma.ai/chestnut/#ready-to-drive \"Permalink\")\n\nchestnut comes in a convenient ready to drive kit, which includes chestnut, an AMD Radeon RX 9060 8GB, and everything needed to power, connect, and mount it in your car.\n\n\n\nThe GPU mounts out of the way in the passenger footwell or under the passenger seat. Run the included cable to comma four, connect to your car’s cigarette lighter, and you’re ready to run openpilot’s biggest models.\n\n## Get your chestnut [Permalink](https://blog.comma.ai/chestnut/#get-your-chestnut \"Permalink\")\n\nchestnut is available today in the [comma shop](https://comma.ai/shop/chestnut).\n\n- chestnut ready to drive is available now for $799.\n- tiny chestnut is available now for $249 for use with your own GPU.\n- Both versions come with a 30-day money-back trial and a one-year hardware warranty.\n\nExcited about scaling up? Join us in the **#hw-chestnut** channel on [Discord](https://discord.comma.ai)!\n\n*Want to help scale up openpilot? We’re [hiring](https://comma.ai/jobs) for autonomy, product, and operations.*\n", "dateCreated": "2026-09-19T21:30:23.395188495+00:00", "date_created": "2026-08-12T21:00:00+00:00", "date_modified": "2026-08-12T21:00:00+00:00", "date_published": "2026-08-12T21:00:00+00:00", "description": "Upgrade your comma four’s compute to run openpilot’s biggest models. Today, we’re launching chestnut, a compute upgrade for comma four. Want to run big models? Nothing beats a desktop GPU. We’re also launching the first chestnut-class…", "external_url": "https://blog.comma.ai/chestnut/", "git": { "commit": "ebb0941c5c06a2ab2377342d607535262a3e969e", "path": "items/blog-comma-ai/2026/08/2026-08-12-introducing-chestnut.md", "permalink": "https://github.com/aymericbeaumet/aggr-instance/blob/ebb0941c5c06a2ab2377342d607535262a3e969e/items/blog-comma-ai/2026/08/2026-08-12-introducing-chestnut.md" }, "id": "urn:aggr:item:2b9dfe8a24749e519842bf3b3be30e97afaf17ff", "isBasedOn": { "@id": "https://blog.comma.ai/chestnut/" }, "mainEntity": { "@id": "https://blog.comma.ai/chestnut/", "@type": "CreativeWork", "archivedAt": { "@id": "https://aggr.aymericbeaumet.com/items/blog-comma-ai/2026-08-12-introducing-chestnut/#webpage" }, "dateModified": "2026-08-12T21:00:00+00:00", "datePublished": "2026-08-12T21:00:00+00:00", "headline": "Introducing chestnut", "timeRequired": "PT2M", "url": "https://blog.comma.ai/chestnut/", "wordCount": 433 }, "name": "Archived snapshot: Introducing chestnut", "preview": null, "reading_minutes": 2, "replicated_at": null, "source": "blog-comma-ai", "source_name": "comma.ai blog", "summary": "Upgrade your comma four’s compute to run openpilot’s biggest models. Today, we’re launching chestnut, a compute upgrade for comma four. Want to run big models? Nothing beats a desktop GPU. We’re also launching the first chestnut-class model, which is …", "tags": [], "temporalCoverage": "2026-09-19T21:30:23.395188495+00:00", "title": "Introducing chestnut", "url": "https://aggr.aymericbeaumet.com/items/blog-comma-ai/2026-08-12-introducing-chestnut/", "word_count": 433 }