Blend Labs Q2 Revenue and Profitability Rise on Software Growth

Blend Labs posted second‑quarter 2026 revenue of $33.8 million, a 7% increase from a year ago, as software platform sales grew 7% to $31.4 million. The company’s GAAP operating loss shrank to $1.6 million from $4.8 million in the same period last year, while non‑GAAP operating income climbed to $7 million from $4.6 million. Non‑GAAP gross margin improved to 78% and the firm ended the quarter with $44.9 million in cash and no debt.

Co‑founder and head of Blend Nima Ghamsari called the quarter “disciplined” and “profitable,” and highlighted the commercial launch of Autopilot – an AI‑powered tool that automates loan fulfillment tasks. The company added or expanded 14 customer contracts in Q2, including six deals involving Autopilot, which officially launched on July 1. Blend also repurchased 11 million shares for $18.2 million during the quarter.

Looking ahead, Blend guided for Q3 revenue of $31.5 million to $33.5 million and non‑GAAP operating income of $3.5 million to $4.5 million. For Q4, the company adopted a cautious stance, citing a “higher for longer” interest‑rate environment that could dampen mortgage refinancing – it expects Q4 market volumes to drop about 11% year over year and its own funded‑loan volumes to decline 10% to 15%.

Inside Blend’s AI‑Powered Pivot and the Mortgage Tech Landscape

The Push into AI Agents

The most consequential part of Blend’s quarter was the rollout of “Blend 3.0,” which Ghamsari described as an agentic‑first architecture. He reported that the engineering team’s throughput has risen 3.6 times since January with roughly the same headcount, because AI agents take a first pass on tasks such as code changes. Ghamsari suggested that throughput could be 10 times by year‑end.

Advertisement

On the product side, early Autopilot data indicates it automates an average of four and a half hours of fulfillment tasks per loan. Blend claims that its pilot lenders are seeing a 10% to 15% improvement in pull‑through rates and two to four days of cycle‑time reduction. Six lenders had signed on by July 1, including Onity Mortgage Corp.

Navigating a ‘Higher for Longer’ Rate Environment

Blend’s cautious Q4 guidance highlights a core vulnerability: mortgage‑origination volumes remain sensitive to interest rates. With rates expected to stay elevated, refinance activity is likely to remain muted, pressuring total market volumes. Blend’s own forecast of a 10%–15% decline in funded‑loan volumes underscores this exposure, though the company’s growing software‑platform revenue and cost discipline provide a buffer.

What the Numbers Reveal

The Q2 results show a company that is gradually shifting its revenue mix toward higher‑margin software while keeping expenses in check. Non‑GAAP operating income more than doubled year over year, and the balance sheet is debt‑free with $44.9 million in cash. The share buyback signals management’s confidence in sustainable cash generation, even as the macro environment remains uncertain.

What Blend’s AI Gains Mean for Lenders and Investors

  • Lenders considering Autopilot: early adopters are reporting a 10‑15% lift in pull‑through rates and a reduction of 2–4 days per loan cycle. The product became commercially available on July 1; you can now evaluate it with real loan files.
  • With Blend projecting Q4 funded‑loan volumes to drop 10‑15%, lenders who sign multi‑year software deals now may lock in terms before a potential volume contraction tightens budgets.
  • Investors should track third‑quarter results against Blend’s own guidance – revenue of $31.5 million to $33.5 million and non‑GAAP operating income of $3.5 million to $4.5 million. A miss could indicate slower‑than‑expected Autopilot adoption.
  • The share buyback of 11‑million shares for $18.2 million leaves $13.2 million remaining under the current authorization. Further repurchases would signal that management sees the stock as undervalued, particularly if Q4 volume fears prove overblown.

Risk & Opportunity Assessment

Commercial RiskMediumBlend’s revenue is tied to mortgage origination volumes, and the Q4 outlook assumes an 11% drop in market volumes in a ‘higher for longer’ rate environment. However, the company’s shift toward software revenue and a cash‑rich, debt‑free balance sheet provide cushion.
Competitive RiskMediumAutopilot’s early efficiency gains (4.5 hours saved per loan, 10–15% conversion lift) create a first‑mover advantage, but larger fintech rivals could develop similar AI agents. Blend’s 3.6x engineering throughput increase with AI shows it is moving quickly internally.
Regulatory RiskLowNo new regulatory threats were raised in this earnings report, and mortgage technology providers face standard compliance requirements that appear manageable.
Reputation RiskLowAutopilot is an early‑stage product, and if it fails to deliver promised cycle‑time or conversion improvements consistently, customer trust could erode. However, current pilot data is positive and customer sign‑ups are growing.
Technology DisruptionHighBlend 3.0’s agentic architecture and internal AI agents that already handle code changes represent a fundamental shift in how mortgage technology firms operate. If successful, this could disrupt incumbent, non‑AI workflows and raise the bar for the entire industry.
Commercial OpportunityHighAutopilot’s automation of 4.5 hours per loan and a 10–15% conversion improvement could drive significant demand from lenders seeking to cut fulfillment costs. As of July 1, six lenders had already signed, suggesting a strong pipeline and potential for rapid revenue growth.