Multisig accounts API

Sep 19, 2014 • 1 分鐘閱讀

Today we’re announcing a new developer tool in beta: our multi-sig API.

We’re still working on a great consumer interface around this, but for today we’re releasing it to developers only.

Does this put me in control of my Bitcoin?

Yes. When you go and create a multisig account, you provide the extended (BIP32) public keys, which only enable us to create new receive addresses for your convenience. Only you have the extended private keys that can sign transactions and move bitcoin out of these addresses.

How to use

Learn how to use the new API in this tutorial.

Multisig accounts are considered an advanced feature and our API is currently in beta, so only use small amounts of bitcoin for the moment.

最近的故事

Singapore Sees 43% Drop in Crypto Scam Losses and Partnerships Like Ours Are Helping Drive the Change

Coinbase

Oct 13, 2026•2 分鐘閱讀

TL;DR: Singapore’s crypto scam losses fell 43% year over year, from S$114.9 million to S$65.5 million, as a result of stronger enforcement, public education, and partnerships between the Anti-Scam Centre and industry leaders like Coinbase. Through four joint operations, Coinbase’s Global Intelligence team helped identify 720+ victims and prevent more than S$18.9 million in potential losses, while a new anti-scam awareness campaign reached Singapore retail customers.

International

Part 3/3 Owning Intelligence Series: How Coinbase Built, Evaluated, and Post-Trained a Fraud Agent

Yao Ma, Xuwei Tan, Akshit Trehan, Aman Choudhary, Anyscale Team: Eric Tang, Sumanth Hegde, Seiji Eicher, Jess Kong

Oct 8, 2026

TL;DR: Using SkyRL on Anyscale, we post-trained Qwen3.5-9B for Onramp fraud detection and served it with Ray Serve LLM and vLLM. On our benchmark, it exceeded Opus 4.5 across all four detection metrics, including 9.6% higher F1. In online inference, we reduced end to end latency by 55%.

Engineering

Part 2/3 Owning Intelligence Series: How Coinbase Built, Evaluated, and Post-Trained a Fraud Agent

Yao Ma, Xuwei Tan

Oct 7, 2026

TL;DR: Newer LLMs are not necessarily better at domain-specific tasks. On our fraud benchmark, newer Opus, Sonnet, and GPT versions all had lower recall, F1, and dollar-weighted recall—even when precision improved. The lesson: choose models using high-quality, domain-specific benchmarks, with reproducible evaluation and ideally deterministic scoring against verified outcomes. These findings also motivated our next step: post-training open-weight models for greater control over quality, cost, and deployment.

Engineering

免責聲明：Coinbase Financial Services Europe Ltd. 會讓符合相應資格的 EEA 客戶透過 Coinbase Advanced 平台來進行期貨交易。（CySEC 授權 374/19）。想要交易衍生性商品的話，客戶們會需要先通過我們的標準評估檢查，好讓我們藉此確定對方是否適合交易這項產品。