SoundHound AI: The Growth Is Real, but Don't Confuse a Squeeze for a Setup

SoundHound AI reported record Q2 2026 revenue of $61.9 million on August 5 — a 45% jump year-over-year and a clear beat on consensus estimates of $52.4 million. Management raised the full-year revenue outlook to $230 million to $260 million, up from the prior midpoint of $242.5 million. The stock surged 26% in premarket trading the next day, then continued climbing. Over five trading days, SOUN is up more than 30%.

The headline narrative is straightforward: enterprise AI demand is fueling growth, and SoundHound is finally proving it. And the revenue numbers genuinely support that — $61.9 million is 10 times what the company was pulling in when it went public in Q2 2022.

But here's the part the headline doesn't tell you. Before earnings, approximately 42% of SoundHound's public float — roughly 169 million shares — was sold short. That's not mild skepticism; that's one of the highest short interests in the market. When the earnings beat landed, those shorts didn't just stand down. They covered, and that forced buying amplified the move far beyond what the revenue number alone would have produced.

The squeeze explains the magnitude of the rally, not the direction. The direction comes from the fundamentals. Revenue growth is real. Gross margins improved to 45.1% on a GAAP basis in Q2, up from 39.0% a year earlier. Non-GAAP gross margins held at 58.4%, up 8 percentage points from Q1. Adjusted EBITDA — earnings before interest, taxes, depreciation, and amortization, a rough proxy for operating cash generation — narrowed its loss 33% year-over-year to $9.6 million.

The OASYS agentic AI platform, launched in May, appears to be shortening enterprise sales cycles and accelerating deployments. Management cited an eight-figure customer commitment that went from demo to contract in under 90 days. New wins and renewals span healthcare, banking, automotive, and restaurants — including multi-year partnerships in Latin America and expanding deployments with Stellantis and Hyundai.

AInvest's aggregate signal labels the stock a Buy, with high marks across fundamentals and liquidity. That's consistent with the earnings data. The growth trajectory is legitimate.

So what's the setup now?

The numbers that matter for risk/reward sit behind the growth story. SoundHound burned $60 million in operating cash during the first half of 2026, against $203 million in cash at quarter end and no debt. At the current burn rate, that's roughly four quarters of runway. The company partially offset the burn by raising $48.5 million through Class A share sales in H1, diluting existing shareholders by about 3.4%. Outstanding shares rose from 390.1 million to 403.3 million.

The forward valuation already prices in a best-case scenario. At a forward P/E of 31.8 times, the market is assuming SoundHound reaches meaningful profitability in 2027. That requires two big things to work: the LivePerson acquisition closes by year-end (as expected), and the combined company delivers on management's claim of $350 million to $400 million in 2027 revenue with gross margins expanding above 70%. Management believes the combined footprint — 25 of the Fortune 100 — will drive an accelerated path to profitability. That's the thesis the forward multiple is built on.

The problem isn't the thesis. It's the price you're paying for it after a 30% weekly squeeze. At a trailing price-to-sales ratio of 19 times, SoundHound trades well above the broader technology sector median of about 6.5 times. It is below its own five-year P/S median of 14.2 times, which is a defensible data point — but only if you believe the 80% revenue growth rate is sustainable through the LivePerson integration, which carries genuine execution risk. Combining two enterprise platforms, two engineering teams, and two customer bases is not a mechanical process.

On the price-action side, the stock's RSI sits at 69 — approaching overbought territory. The 50-day moving average is $6.87, which the stock cleared decisively. But the 200-day moving average is $9.20, and SOUN hasn't reclaimed it yet. Year-to-date, the stock is down 19.6%. Over the rolling 12 months, it's down 41.2%. The longer-term structure hasn't confirmed a bottom despite the earnings-driven pop.

That tells me this is not a chasing setup. The growth story is real, the moat is building through OASYS, and the market's prior skepticism — 42% short interest — was arguably overdone. But the post-earnings mechanics have already delivered most of the immediate upside. The forward multiple assumes a flawless LivePerson execution, and the cash burn timeline means any delay or integration stumble could force more dilution.

I don't think investors need to chase the rally here. The underlying thesis — accelerating enterprise AI adoption, a platform that's shortening sales cycles, and a company whose revenue has grown 10x since going public — has merit. But the better risk/reward for new positions is on a pullback toward that 50-day moving average or below, where the squeeze premium has dissipated. For those already positioned, holding makes sense if you believe in the LivePerson synergy case and can tolerate continued dilution risk.

I would reassess the bullish case if LivePerson hits regulatory or integration delays, if cash burn accelerates beyond the current $60 million per half-year pace, or if the stock fails to hold above its 50-day moving average once selling pressure returns. Until then, the growth is the growth — just don't mistake a short squeeze for a buying signal.