Canaan’s first-quarter earnings painted a mixed picture. Revenue fell sharply to $62.7 million, down from $196.3 million in Q4 2025 and below the $82.8 million reported a year earlier. The net loss widened to $88.7 million, while adjusted EBITDA losses nearly doubled to $76.3 million.
At the same time, the company’s crypto treasury reached a record high. As of March 31, Canaan held 1,807.60 Bitcoin and 3,951.53 Ethereum. At current market prices, that stash is worth roughly $148 million. That creates an interesting tension inside the quarter: a weakening hardware business paired with a growing digital asset balance sheet.
The hardware cycle is the pressure point
ASIC miner sales fell to $42.9 million from $164.9 million in the previous quarter. Canaan attributed the decline to lower computing power sold and a lower average selling price, which it linked to weaker market demand after Bitcoin’s price drop.
That comment matters because ASIC makers depend on miner confidence. When miners expect new machines to pay for themselves quickly, orders flow. When power costs, difficulty, or hashprice squeeze margins, demand dries up fast. Q4 had benefited from a large U.S. customer order, which made the sequential drop look sharper than it might otherwise be. But the Q1 language still pointed to a broader problem: weaker unit demand and lower pricing.
Outside Canaan, miner economics were showing signs of recovery. Hashrate Index noted average USD hashprice rose 8.5% to $33.92 per PH per day in April after two all-time-low monthly averages. Even with hashprice near $40 in early May, marginal hashrate had not returned to the network. That suggests miners are still cautious, which does not bode well for new hardware orders.
The treasury is the counterweight
Canaan’s Bitcoin and Ethereum holdings continued to climb. The company said it converted stablecoin proceeds from miner sales into Bitcoin during January, pushing its reserve to 1,778 BTC and 3,951 ETH by the end of that month. By March 31, the balance reached 1,807.60 BTC and 3,951.53 ETH. After the quarter closed, Canaan added 90 BTC from self-mining and 3 BTC from customer payments in April.
That mechanism shifts how the quarter reads. Canaan’s crypto balance now reflects ongoing operating decisions alongside legacy holdings. Some miner-sale proceeds flowed into Bitcoin, and self-mining kept adding BTC even as mining revenue fell from Q4. The company now has elements of three different business models: a pure ASIC supplier, a miner, and a treasury holder. That makes its reported weakness harder to interpret through a single lens.
The operating loss remains the counterpoint. Canaan guided Q2 revenue to only $35 million to $45 million, below the already weak Q1 result. That guidance means the balance sheet may become a larger part of the narrative precisely because the income statement is not yet showing recovery.
Infrastructure gives Canaan a third lane
Canaan’s Q1 release also pushed a broader infrastructure message. The company highlighted its Nordic hash-to-heat deployment and a stake in West Texas ABC Projects, which sits closer to energy and compute infrastructure than traditional machine sales. Those details belong behind the core numbers, but they help explain why Canaan is looking beyond the next ASIC order cycle.
Public miners have already been pulled toward energy, hosting, and AI or high-performance compute strategies as mining margins tighten. Canaan’s version is different because it sits upstream. It sells into miners, operates its own mining exposure, holds a growing crypto stack, and is testing energy-linked infrastructure projects.
That mix can help if hardware demand stays weak, but it also makes the investment story more complicated. A buyer of Canaan’s stock is reading ASIC sales, Bitcoin price exposure, self-mining output, and management’s ability to turn infrastructure projects into durable revenue. That complexity is why the quarter stops being a basic miss-versus-expectations story.
The next test is straightforward: whether Q2 revenue and product pricing stabilize enough to make Q1 look like a weak transition quarter, or whether Canaan’s guided decline pushes the story further toward treasury, self-mining, and infrastructure exposure. If customer demand improves, Canaan can still be read primarily as a cyclical ASIC supplier with a growing Bitcoin and Ethereum balance. If revenue follows guidance lower and the crypto stack keeps rising, the market will have more reason to treat the company as a hybrid: part hardware seller, part miner, part Bitcoin treasury, and part energy-compute operator.
For now, the sourced record supports the tension rather than a clean verdict. Q1 showed a weaker hardware business, a wider loss, lower mining revenue, and a larger crypto treasury. That combination makes Canaan one of the clearer examples of how the Bitcoin mining trade is changing: even the company selling the picks and shovels is increasingly carrying the asset risk its customers face every day.
