# Anthropic in talks with Samsung to build custom Claude inference chips at 2nm

> Anthropic is in early talks with Samsung Electronics to manufacture its own custom AI inference chips on a 2-nanometer process, according to reports. The goal is straightforward economics: the company is reportedly facing a compute bill north of $1.25 billion a month, and custom silicon tuned specifically for running Claude could cut costs while reducing dependence on Nvidia's GPUs. The move follows OpenAI's in-house Jalapeño chip, which reportedly delivered around 50 percent cost savings versus GPU inference. Anthropic's talks are still preliminary, with no decisions yet on chip specs or how the parts would slot into servers, and existing options like AWS Trainium, Google TPUs, and Nvidia GPUs would remain central to its strategy. Every major lab is now discovering that the cost of serving models, not training them, is the line item that scales without limit. Designing custom inference chips is expensive and slow, but at Anthropic's spend even a partial cost cut pays for the effort quickly, and it loosens Nvidia's grip on the whole industry.

_Section: [Daily AI Updates](https://www.wortins.com/daily-ai) · Source: TechCrunch · Published Friday, July 17, 2026_

## Wortins' read

Anthropic is in early talks with Samsung Electronics to manufacture its own custom AI inference chips on a 2-nanometer process, according to reports. The goal is straightforward economics: the company is reportedly facing a compute bill north of $1.25 billion a month, and custom silicon tuned specifically for running Claude could cut costs while reducing dependence on Nvidia's GPUs. The move follows OpenAI's in-house Jalapeño chip, which reportedly delivered around 50 percent cost savings versus GPU inference. Anthropic's talks are still preliminary, with no decisions yet on chip specs or how the parts would slot into servers, and existing options like AWS Trainium, Google TPUs, and Nvidia GPUs would remain central to its strategy. Every major lab is now discovering that the cost of serving models, not training them, is the line item that scales without limit. Designing custom inference chips is expensive and slow, but at Anthropic's spend even a partial cost cut pays for the effort quickly, and it loosens Nvidia's grip on the whole industry.

## Source

[Read the full story at TechCrunch](https://techcrunch.com/2026/07/02/anthropic-is-discussing-a-new-custom-chip-with-samsung/)

## Related coverage

- [Clearview AI Is Testing an AI Tool That Would Let Cops Unearth Your Life Online](https://www.wortins.com/story/clearview-ai-is-testing-an-ai-tool-that-would-let-cops-unear-8c573d0e) — [Wired](https://www.wired.com/story/clearview-ai-is-testing-an-ai-tool-that-lets-cops-instantly-unearth-your-online-activity/)
- [Trump officials say AI will help save rural health care. Some leaders in the field don’t believe it](https://www.wortins.com/story/trump-officials-say-ai-will-help-save-rural-health-care-some-bf0e7d83) — [STAT](https://www.statnews.com/2026/09/10/rural-health-care-ai-adoption-challenges-part-4-unraveled-series/?utm_campaign=rss)
- [Farmers Embrace AI More Than Any Other Tech, McKinsey Says](https://www.wortins.com/story/farmers-embrace-ai-more-than-any-other-tech-mckinsey-says-78dfb3ec) — [Insurance Journal](https://www.insurancejournal.com/news/national/2026/09/09/884469.htm)
- [Sequoia doubles down on Cymphony as AI agents create new enterprise security risks](https://www.wortins.com/story/sequoia-doubles-down-on-cymphony-as-ai-agents-create-new-ent-e2ff7229) — [TechCrunch](https://techcrunch.com/2026/09/09/sequoia-doubles-down-on-cymphony-as-ai-agents-create-new-enterprise-security-risks/)
- [Powering AI is an architecture problem](https://www.wortins.com/story/powering-ai-is-an-architecture-problem-15281f16) — [MIT Technology Review](https://www.technologyreview.com/2026/09/10/1141649/powering-ai-is-an-architecture-problem/)
- [Top AI spenders cut per-employee costs by nearly 10 percent in August](https://www.wortins.com/story/top-ai-spenders-cut-per-employee-costs-by-nearly-10-percent--b31ffb0b) — [The Decoder](https://the-decoder.com/top-ai-spenders-cut-per-employee-costs-by-nearly-10-percent-in-august/)

---

_Curated and written by [Wortins](https://www.wortins.com) — The daily AI briefing. Every story links to its original source; the "Wortins read" on each is our own original analysis. [About Wortins & our editorial approach](https://www.wortins.com/about)._