SoundHound AI, Inc.
SOUN is growing fast, but at 12.5x sales with still deeply negative margins and a missed profitability promise, the stock price assumes execution that the numbers do not yet support.
SOUN's problem is not whether demand exists. Demand plainly exists. FY2025 revenue reached $0.2B, up 84.3%, Q2 FY2026 set a company record at $61.9M, and management felt confident enough to raise FY2026 revenue guidance to $230M-$260M. The company is also improving its delivery stack, with non-GAAP gross margin up to 58% and in-house models lowering infrastructure cost. Those are the facts behind the bull case, and they matter.
The issue is what investors are paying for that progress. At $6.14, SOUN carries a $2.6B market cap and trades at 12.5x sales, versus a peer average of 5.4x. That premium might be justified if operating leverage were already showing up clearly. It is not. FY2025 operating margin was -185.0% and net margin -184.3%. Worse, the latest quarter's adjusted EBITDA margin moved the wrong way sequentially, falling to -58.0% from -32.1% in Q1, even as revenue hit a record. Investors are still underwriting a future margin structure, not a present one.
The second problem is credibility on the path to profitability. In the 2024 Q3 call, management explicitly said it expected adjusted EBITDA profitability by the end of 2025. Instead, Q2 FY2026 still showed -$9.6M of adjusted EBITDA. Revenue execution has beaten expectations, but profitability execution has missed. That matters because the next leg of the story leans heavily on management's claim that FY2027 revenue could reach $350M-$400M with LivePerson. That target may be achievable, but it is partly acquisition-dependent and the FY2026 guide explicitly excludes any LivePerson contribution due to closing uncertainty.
The stock is also no longer washed out enough to erase those issues. Yes, it is down 72.3% from the 52-week high and sits 23.9% below the 200-day average, which reduces some speculative excess. But extreme short interest at 39.89% cuts both ways. It can fuel squeezes, yet it also signals that a large part of the market still questions the economics. The case for owning SOUN here depends on three things: gross margin pushing toward management's above 70% goal, LivePerson closing and integrating cleanly, and EBITDA losses narrowing fast enough to justify a premium sales multiple. Until those are visible in reported numbers, the stock still looks expensive.
FY2026 exits with stronger gross margin, LivePerson closes cleanly, FY2027 revenue tracks toward $400M, and heavy short interest drives an overshoot above fundamentals.
Growth continues but profitability remains inconsistent, leaving valuation to converge toward the engine's sales_quality_blend target as enthusiasm fades.
LivePerson slips or underdelivers, revenue lands near the low end of $230M-$260M, margins stay deeply negative, and the stock compresses toward a premium but more defensible software-sales multiple.
Quant call (SELL) diverges from analyst consensus (Strong Buy: 2 buy / 1 hold / 0 sell) — resolve explicitly.
We don't buy stocks where they are, we buy them where they're going.
Forward earnings power remains the core problem. There is no useful trailing P/E because earnings are still negative, and even on management's FY2027 revenue ambition of $350M-$400M including LivePerson, the current $2.6B market cap implies roughly 6.5x to 7.4x forward sales on a deal-adjusted number that is not yet closed. The moat is better than two years ago because OASYS adoption, channel-led wins across 20 countries, and more in-house models can lift gross margin, but over the next 3-5 years voice AI becomes more crowded as foundation-model vendors and larger platforms push into enterprise workflows. AI is both tailwind and threat here: it expands TAM in automotive, customer service, and voice commerce, but it also lowers barriers in application-layer software unless SoundHound proves durable distribution and profitable deployment.
- 12.5x P/S versus 5.4x peer average leaves little room for another margin miss or slower FY2027 conversion.
- FY2026 revenue guidance of $230M-$260M implies growth, but the latest quarter still posted adjusted EBITDA margin of -58.0%.
- Management now talks to $350M-$400M FY2027 revenue with LivePerson, yet FY2025's promised year-end EBITDA profitability was not achieved.
- FY2025 revenue reached $0.2B, up 84.3% YoY, showing real demand rather than a purely narrative-driven AI trade.
- Q2 FY2026 revenue was a record $61.9M, up 45% YoY, and management raised FY2026 revenue guidance to $230M-$260M after the first half.
- Non-GAAP gross margin improved to 58% in Q2 FY2026, up 800 bps sequentially, helped by cloud optimization and more in-house models.
- The balance sheet is a real support, with $203M cash, no debt, debt/equity 0.01, and a 3.91 current ratio, which lowers financing risk while the company scales.
- FY2025 operating margin was still -185.0% and net margin -184.3%, even after improving from a prior -403.0% operating margin.
- Q2 FY2026 adjusted EBITDA was -$9.6M and the margin worsened to -58.0% from -32.1% in Q1, showing that growth is not yet translating cleanly into profitability.
- The stock trades at 12.5x sales, more than 2.3x the 5.4x peer average, while comparables themselves are noisy, with peer P/E dispersion at 6.3x.
- Execution credibility is mixed because management said in 2024 Q3, "In accordance, we expect to achieve adjusted EBITDA profitability by the end of 2025," but the latest reported Q2 FY2026 still showed negative 58.0% adjusted EBITDA margin.
| Metric | Value | Context |
|---|---|---|
| Current price | $6.14 | Reference point for scenario analysis and engine target comparison |
| Quant price target | $3.54 | -42.3% vs current, Low confidence, basis: sales_quality_blend |
| Method blend | P/S Multiple $2.43, Quality-adjusted $6.04 | Wide spread shows valuation sensitivity to assumptions |
| FY2025 revenue | $0.2B | +84.3% YoY, strong growth but still small absolute scale |
| FY2025 EPS | $-0.03 | -213.2% YoY, profitability remains weak |
| FY2025 operating margin | -185.0% | Improved from prior FY -403.0%, but still deeply negative |
| Valuation | P/S 12.5 | Versus peer average 5.4, premium despite losses |
| Balance sheet | Debt/Equity 0.01, Current ratio 3.91 | Healthy liquidity and minimal leverage |
| Q2 FY2026 revenue | $61.9M | Record quarter, +45% YoY, above management expectations |
| Q2 FY2026 adjusted EBITDA margin | -58.0% | Down 2,590 bps sequentially from Q1's -32.1% |
| FY2026 revenue guidance | $230M-$260M | Raised after strong first half, excludes LivePerson contribution |
| 52-week position | $5.65-$22.17 | Current price is -72.3% vs high and -23.9% below 200-day average |
The biggest risk to a bearish stance is that revenue scales faster than expected into management's $350M-$400M FY2027 framework, allowing gross margin above 70% and making today's 12.5x sales look less demanding.
The next catalyst is November 5, 2026 earnings, where investors need proof that raised FY2026 revenue guidance can hold while adjusted EBITDA improves from -58.0% and LivePerson closing timing becomes clearer.
Positioning is split between institutional sponsorship and speculative pressure. The top three institutions own 39.9%, but insider ownership is only 1.9%, and short interest is extremely high at 39.89%, or 156.9M shares with 5.5 days to cover. Options positioning is mildly supportive because the near-term regime is positive and zero-gamma sits at $5.53, below spot, but that can fuel violent squeezes without changing intrinsic value.
Management said enterprise AI was the largest contributor in Q2 FY2026, and in the 2024 Q3 call said AI agent customer service was already more than half of business.
Recent commentary points to Asia growth extending to 5 quarters, but the company did not provide a revenue split by geography in the supplied data.
Q2 FY2026 growth was described as broad-based across these verticals, with automotive and enterprise AI still important demand vectors.
Record Q2 and a raised FY26 revenue outlook driven by OASYS-led demand.
Management has delivered on growth better than on profitability. In the 2024 Q3 call, the company guided FY2025 revenue to $155M-$175M and FY2025 actual revenue reached $0.2B, so revenue execution beat that framework. But management also said, "In accordance, we expect to achieve adjusted EBITDA profitability by the end of 2025," and that clearly did not happen, with Q2 FY2026 adjusted EBITDA still -$9.6M and margin -58.0%. Product and cost execution is mixed but improving: gross margin moved from 49% in the 2024 Q3 period to 58% non-GAAP in Q2 FY2026, Polaris and in-house model deployment did progress, and the earnings beat record has been strong in 7 of the last 8 reported quarters, yet the large 2025-02-27 EPS miss of -527.3% against consensus shows forecasting remains uneven.
This was not only a record Q2, but our largest all-time quarter with $62 million in revenue.
With the strong first half of the year, we are increasing our revenue expectations to be in the range of $230 million to $260 million.
The guidance we provided does not include the acquisition. And because of the uncertainty in the timing of the closing, we are not able to provide clear guidance on what it would look like.
we talked about that last time was at minimum, the range would be $350 million to $400 million next year.
Our pipeline has never been this big and our win rate has never been this good.
The moat is real but not yet proven durable. SoundHound has domain-specific voice AI, embedded automotive relationships, and a growing stack of in-house models that improved non-GAAP gross margin to 58% in Q2 FY2026, with management targeting above 70% over time. The concern is durability: application-layer AI is seeing rapid feature convergence, so long-term returns will depend less on raw model quality and more on distribution, workflow integration, and whether products like voice commerce create sticky transaction rails.
Capital allocation has been aggressive on growth and M&A rather than shareholder returns, which fits the stage of the business. Buybacks, dividends, and capex were all $0.0B, while FCF was effectively $0.0B and the company is relying on its balance sheet and stock-market support to fund expansion. The positive is balance-sheet flexibility, with $203M cash and no debt; the risk is that acquisition integration, including LivePerson, stretches operating discipline before the core business turns sustainably profitable.
SOUN rates a SELL. The company has real growth, a strong balance sheet, and genuine AI demand, but the stock still prices in a cleaner path to profitable scale than management has delivered. The $3.54 quant target looks directionally right because today's 12.5x sales multiple is hard to defend against -185.0% operating margin and an acquisition-dependent FY2027 story. A change in view would require sustained EBITDA improvement, not just revenue beats, plus evidence that LivePerson closes and boosts economics without another reset. Until then, the risk is still to valuation, not to the narrative.
≣ Full research note tap to expand
Executive Summary
SELL, Medium conviction. SOUN rates a SELL at Medium conviction. Revenue growth is real, FY2025 revenue reached $0.2B, up 84.3%, and FY2026 guidance was raised to $230M-$260M, but the stock still trades at 12.5x sales versus peers at 5.4x while operating margin remains -185.0% and adjusted EBITDA in the latest quarter was -58.0%. After a -72.3% drawdown from the 52-week high, the setup is less euphoric than before, but valuation still looks ahead of fundamentals.
Investment Thesis
SOUN's problem is not whether demand exists. Demand plainly exists. FY2025 revenue reached $0.2B, up 84.3%, Q2 FY2026 set a company record at $61.9M, and management felt confident enough to raise FY2026 revenue guidance to $230M-$260M. The company is also improving its delivery stack, with non-GAAP gross margin up to 58% and in-house models lowering infrastructure cost. Those are the facts behind the bull case, and they matter.
The issue is what investors are paying for that progress. At $6.14, SOUN carries a $2.6B market cap and trades at 12.5x sales, versus a peer average of 5.4x. That premium might be justified if operating leverage were already showing up clearly. It is not. FY2025 operating margin was -185.0% and net margin -184.3%. Worse, the latest quarter's adjusted EBITDA margin moved the wrong way sequentially, falling to -58.0% from -32.1% in Q1, even as revenue hit a record. Investors are still underwriting a future margin structure, not a present one.
The second problem is credibility on the path to profitability. In the 2024 Q3 call, management explicitly said it expected adjusted EBITDA profitability by the end of 2025. Instead, Q2 FY2026 still showed -$9.6M of adjusted EBITDA. Revenue execution has beaten expectations, but profitability execution has missed. That matters because the next leg of the story leans heavily on management's claim that FY2027 revenue could reach $350M-$400M with LivePerson. That target may be achievable, but it is partly acquisition-dependent and the FY2026 guide explicitly excludes any LivePerson contribution due to closing uncertainty.
The stock is also no longer washed out enough to erase those issues. Yes, it is down 72.3% from the 52-week high and sits 23.9% below the 200-day average, which reduces some speculative excess. But extreme short interest at 39.89% cuts both ways. It can fuel squeezes, yet it also signals that a large part of the market still questions the economics. The case for owning SOUN here depends on three things: gross margin pushing toward management's above 70% goal, LivePerson closing and integrating cleanly, and EBITDA losses narrowing fast enough to justify a premium sales multiple. Until those are visible in reported numbers, the stock still looks expensive.
Key Metrics
| Metric | Value | Context |
|---|---|---|
| Current price | $6.14 | Reference point for scenario analysis and engine target comparison |
| Quant price target | $3.54 | -42.3% vs current, Low confidence, basis: sales_quality_blend |
| Method blend | P/S Multiple $2.43, Quality-adjusted $6.04 | Wide spread shows valuation sensitivity to assumptions |
| FY2025 revenue | $0.2B | +84.3% YoY, strong growth but still small absolute scale |
| FY2025 EPS | $-0.03 | -213.2% YoY, profitability remains weak |
| FY2025 operating margin | -185.0% | Improved from prior FY -403.0%, but still deeply negative |
| Valuation | P/S 12.5 | Versus peer average 5.4, premium despite losses |
| Balance sheet | Debt/Equity 0.01, Current ratio 3.91 | Healthy liquidity and minimal leverage |
| Q2 FY2026 revenue | $61.9M | Record quarter, +45% YoY, above management expectations |
| Q2 FY2026 adjusted EBITDA margin | -58.0% | Down 2,590 bps sequentially from Q1's -32.1% |
| FY2026 revenue guidance | $230M-$260M | Raised after strong first half, excludes LivePerson contribution |
| 52-week position | $5.65-$22.17 | Current price is -72.3% vs high and -23.9% below 200-day average |
Financial Strength
The balance sheet is the best part of the story. SoundHound has no debt, minimal leverage by both debt/equity 0.01 and debt/EBITDA 0.21, a 3.91 current ratio, and management cited $203M cash, so solvency risk is low even with continued losses. The issue is not survival, it is conversion: despite much better scale and some gross-margin progress, operating economics are still far from self-funding, and the sequential deterioration in Q2 adjusted EBITDA margin suggests cost discipline is not yet consistent.
Competitive Position
The moat is real but not yet proven durable. SoundHound has domain-specific voice AI, embedded automotive relationships, and a growing stack of in-house models that improved non-GAAP gross margin to 58% in Q2 FY2026, with management targeting above 70% over time. The concern is durability: application-layer AI is seeing rapid feature convergence, so long-term returns will depend less on raw model quality and more on distribution, workflow integration, and whether products like voice commerce create sticky transaction rails.
Management & Guidance
Management's current revenue guide is clearly stronger than it was six months ago. The company raised FY2026 revenue expectations to $230M-$260M, supported by a record Q2, broader enterprise AI demand, and comments that "Our pipeline has never been this big and our win rate has never been this good." That said, credibility is mixed because the FY2027 framework of $350M-$400M depends in part on a LivePerson close that is still not reflected in FY2026 guidance, and management previously missed its own commitment to reach adjusted EBITDA profitability by the end of 2025. Revenue guidance deserves some credit; profitability guidance deserves skepticism.
Segment Analysis
- Product/segment revenue (Not disclosed): Management said enterprise AI was the largest contributor in Q2 FY2026, and in the 2024 Q3 call said AI agent customer service was already more than half of business.
- Geographic revenue (Not disclosed): Recent commentary points to Asia growth extending to 5 quarters, but the company did not provide a revenue split by geography in the supplied data.
- Automotive, healthcare, financial services, technology (Not disclosed): Q2 FY2026 growth was described as broad-based across these verticals, with automotive and enterprise AI still important demand vectors.
Capital Allocation
Capital allocation has been aggressive on growth and M&A rather than shareholder returns, which fits the stage of the business. Buybacks, dividends, and capex were all $0.0B, while FCF was effectively $0.0B and the company is relying on its balance sheet and stock-market support to fund expansion. The positive is balance-sheet flexibility, with $203M cash and no debt; the risk is that acquisition integration, including LivePerson, stretches operating discipline before the core business turns sustainably profitable.
Management Execution & Track Record
Management has delivered on growth better than on profitability. In the 2024 Q3 call, the company guided FY2025 revenue to $155M-$175M and FY2025 actual revenue reached $0.2B, so revenue execution beat that framework. But management also said, "In accordance, we expect to achieve adjusted EBITDA profitability by the end of 2025," and that clearly did not happen, with Q2 FY2026 adjusted EBITDA still -$9.6M and margin -58.0%. Product and cost execution is mixed but improving: gross margin moved from 49% in the 2024 Q3 period to 58% non-GAAP in Q2 FY2026, Polaris and in-house model deployment did progress, and the earnings beat record has been strong in 7 of the last 8 reported quarters, yet the large 2025-02-27 EPS miss of -527.3% against consensus shows forecasting remains uneven.
Positioning & Flows
Positioning is split between institutional sponsorship and speculative pressure. The top three institutions own 39.9%, but insider ownership is only 1.9%, and short interest is extremely high at 39.89%, or 156.9M shares with 5.5 days to cover. Options positioning is mildly supportive because the near-term regime is positive and zero-gamma sits at $5.53, below spot, but that can fuel violent squeezes without changing intrinsic value.
Bull Case
- FY2025 revenue reached $0.2B, up 84.3% YoY, showing real demand rather than a purely narrative-driven AI trade.
- Q2 FY2026 revenue was a record $61.9M, up 45% YoY, and management raised FY2026 revenue guidance to $230M-$260M after the first half.
- Non-GAAP gross margin improved to 58% in Q2 FY2026, up 800 bps sequentially, helped by cloud optimization and more in-house models.
- The balance sheet is a real support, with $203M cash, no debt, debt/equity 0.01, and a 3.91 current ratio, which lowers financing risk while the company scales.
Bear Case
- FY2025 operating margin was still -185.0% and net margin -184.3%, even after improving from a prior -403.0% operating margin.
- Q2 FY2026 adjusted EBITDA was -$9.6M and the margin worsened to -58.0% from -32.1% in Q1, showing that growth is not yet translating cleanly into profitability.
- The stock trades at 12.5x sales, more than 2.3x the 5.4x peer average, while comparables themselves are noisy, with peer P/E dispersion at 6.3x.
- Execution credibility is mixed because management said in 2024 Q3, "In accordance, we expect to achieve adjusted EBITDA profitability by the end of 2025," but the latest reported Q2 FY2026 still showed negative 58.0% adjusted EBITDA margin.
Valuation & Price Target
Engine price target $3.54 (-42.3% vs current), Low confidence.
- P/S Multiple: $2.43 (45% weight)
- Quality-adjusted: $6.04 (20% weight)
The rating aligns with the quant baseline: SELL, though with somewhat higher conviction than the engine's low-conviction flag because the valuation gap is still stark. The engine's $3.54 target, based on a blend of $2.43 P/S multiple and $6.04 quality-adjusted value, fits a business with a 48/100 quality score, 62/100 risk score, and a stock trading at 12.5x sales versus 5.4x peers while still posting -185.0% operating margin. Sell-side consensus remains Strong Buy from just 3 firms, which is low-signal context and not persuasive against the combination of missed profitability targets, wide peer dispersion of 6.3x, and an acquisition-dependent FY2027 story.
Risk Assessment
Risk score 62/100 (Moderate). The biggest risk to a bearish stance is that revenue scales faster than expected into management's $350M-$400M FY2027 framework, allowing gross margin above 70% and making today's 12.5x sales look less demanding.
The Bottom Line
SOUN rates a SELL. The company has real growth, a strong balance sheet, and genuine AI demand, but the stock still prices in a cleaner path to profitable scale than management has delivered. The $3.54 quant target looks directionally right because today's 12.5x sales multiple is hard to defend against -185.0% operating margin and an acquisition-dependent FY2027 story. A change in view would require sustained EBITDA improvement, not just revenue beats, plus evidence that LivePerson closes and boosts economics without another reset. Until then, the risk is still to valuation, not to the narrative.
Go deeper in the app
Live charts, options analytics, the full earnings-call breakdown, price alerts, and every other stock — updated in real time.
Get Tickr