FUSE Prices an AI Pivot Above a Registered Resale Stack

Summary: Conditional Short Note. Fusemachines is priced at $1.03 after a 22.56% August 4 decline, while an August 4 prospectus supplement covers up to 26.55 million secondary shares, including 6.20 million shares issuable under private placement warrants, against 28.94 million common shares in the latest filed denominator. A fresh Meteora financing adds a senior zero-coupon note due February 12, 2027, a fixed $4.20 conversion price, and a 2.05 million-share warrant. That is a large potential claim stack, not proof of immediate float or an executable short.

Research timestamp: 2026-08-05 14:42:08 Asia/Singapore (UTC+08:00). The latest Nasdaq record available in this run is the August 4 regular-session close. Bid and ask fields were unavailable, so the price is a dated reference only.

Why This Is the Best Opportunity Right Now

The strongest counterargument is decisive: the new $4.20 note and warrant are far out of the money at $1.03, the 26.55 million-share secondary figure includes potential warrant shares, and the filing does not show that holders have sold those shares. Fusemachines could receive the stated $2.05 million purchase price, settle the Forward Purchase Agreement in cash, win a material AI contract, or simply squeeze a crowded low-float tape. Those facts keep this at Short Note quality and make entry.price null.

FUSE still ranks first in this U.S.-only screen because the evidence is unusually fresh and mechanically linked. Three filings arrived at the same time: a resale prospectus, a senior financing with a near-term maturity, and amendments that make a pre-existing Meteora price mechanism more sensitive to the weekly VWAP. The balance sheet adds urgency: $1.774 million of cash against $2.224 million of first-quarter operating cash use, with $19.977 million of total liabilities and a $15.033 million stockholders' deficit. The stock then fell 22.56% on 22.76 million shares, about 18.39 times Nasdaq's indicated average volume.

The disagreement is not “AI is fake.” It is narrower: the market may be treating the operating story as the asset while underweighting the timing, priority, and optionality of the capital-structure claims. That is a high-asymmetry hypothesis, not a confirmed selling event.

U.S.-only opportunity ranking

Rank	Candidate	Latest reference and move	Why it made the screen	Why it lost or won	Screen score
1	FUSE	$1.03, -22.56%, 22.76M volume, 18.39x indicated average	Aug. 4 resale prospectus, Aug. 2 senior note and warrant, weekly-reset Meteora terms, distressed cash conversion	Freshest combination of registered potential supply, maturity pressure, and an observable price dislocation. Execution data is missing.	35/40
2	CYCU	$0.83, -9.29%, 22.54M volume, 1.24x indicated average	Aug. 3 warrant inducement and Aug. 4 acquisition close with cash, secured note, and preferred consideration	Strong liability bridge, but conversion and warrant prices are above spot and management's pro forma forecast is a serious squeeze argument.	31/40
3	NUWE	$1.88, +16.77%, 35.73M volume, 12.60x indicated average	Aug. 3 registered direct at $2.59 with matching private warrants, plus going-concern language	Financing is fresh, but the price is below the financing anchor and preliminary Q2 revenue and new cash can support a sharp rebound.	30/40
4	SXTP	$1.23, +6.03%, 87,777 volume, 0.77x indicated average	July 30 private placement with two warrant series at $1.49	Distress is real, but low volume and a small tape make the execution path weaker than FUSE.	27/40

The market references in this table are Nasdaq last-sale records accessed on August 5, not verified live quotes. Candidate filing sources are listed at the end of the article.

Why This Can Jump Or Dump More Than 5% Soon

The August 4 close already moved 22.56% lower, from a $1.33 previous close to a $1.03 last sale, with a reported session range of $0.9771 to $1.42. A further move of more than 5% is plausible within the next several sessions because the stock is small, the most recent volume was 18.39 times the indicated average, and the market has no verified NBBO or depth record in this research pass.

The short-side trigger would be a new resale, exercise, post-financing share-count, or cash filing followed by failed continuation and a failed bounce. The squeeze trigger would be the reverse: cash receipt, a customer or strategic announcement, FPA cash settlement, no actual resale, or a borrow recall while the price is near the $0.7508 52-week low. The direction is uncertain because the same missing market-structure data that makes a short attractive also makes it dangerous.

Evidence quality is high for the filings and medium for the price path. The exact date and volume are observable. Borrow, short interest, options open interest, dealer flow, fund flow, spread, depth, and exit liquidity are not verified.

What Should Surprise the Reader

The surprising fact is not simply that 26.55 million shares are in a prospectus. It is that the filing places three different capital states beside one another:

Up to 26.55 million secondary shares can be resold, with 6.20 million issuable under private placement warrants.
Up to 9.49 million additional shares can be issued upon public-warrant exercise at $11.50, a price so far above spot that it is a long-dated claim, not current supply.
The new senior note and warrant use a fixed $4.20 price, while the amended Meteora mechanisms refer to a weekly VWAP and a $0.85 minimum for the shortfall warrant.

The latest filed common denominator is 28.94 million shares. The ratio of the secondary prospectus maximum to that denominator is 91.7%, but that arithmetic is not a dilution forecast. Some rows are common shares, some are exercisable claims, and the filing does not prove that any holder is selling today.

The Setup

Fusemachines is an AI services company formed through a post-SPAC combination. The latest balance-sheet information is for March 31, 2026 and was reproduced in the August 4 424B3 prospectus supplement. It reported:

Item	Amount	Interpretation
Revenue, Q1 2026	$1.879M	Down from $1.954M in Q1 2025
Gross profit, Q1 2026	$0.948M	Positive gross profit does not cover operating cost
Operating loss, Q1 2026	$3.035M	The operating model remained cash-intensive
Net cash used in operations, Q1 2026	$2.224M	Latest disclosed quarterly operating cash use
Cash and equivalents, March 31	$1.774M	Less than one quarter of the latest operating cash use, as a ratio only
Total liabilities, March 31	$19.977M	Includes the $8.104M FPA derivative liability
Stockholders' deficit, March 31	$15.033M	Negative book value, not a valuation target
Common shares, March 31 and May 14 filing date	28.939M	Latest filed denominator used for the price-only market-value reference

The cash-to-quarterly-burn ratio is about 0.80. It is not a runway date. The company has receivables, financing commitments, an existing FPA, and possible future capital, while the new note's stated purchase price is not yet confirmed by a subsequent cash-flow statement.

The Market Price

Measure	August 4, 2026 Nasdaq record	Evidence boundary
Last sale	$1.03	Dated regular-session reference, accessed August 5
Change	-$0.30, -22.56%	Versus $1.33 prior close
Day high / low	$1.4200 / $0.9771	The low is not a short entry
Volume	22,755,287	About 18.39x the indicated average volume
Indicated average volume	1,237,061	Nasdaq summary field, provider basis not independently audited
52-week high / low	$4.25 / $0.7508	The stock is close enough to the low for squeeze risk to matter
Filed-share reference value	About $29.8M	28.939M latest filed shares multiplied by $1.03; current share count may differ
Bid / ask	Not available	No NBBO, spread, or depth verification

The market price is therefore a reference level, not an executable entry. A move from $1.03 to the base target of $0.82 would be a 20.4% gross short return before borrow, fees, slippage, taxes, and gap risk. A move to $1.75 would be a 69.9% loss on the same price basis.

The Positioning

The observable positioning is structural, not directional. The filings show potential claims and registration rights. They do not show a live short-interest percentage, borrow availability, cost to borrow, recall terms, dealer gamma, or holder sales.

Capital state	Filed amount	What it does and does not prove
Common shares outstanding	28.939M	Latest reported denominator as of March 31, with the 10-Q cover date May 14. It may predate the August financing.
Secondary resale prospectus	Up to 26.549M shares	Potential resale from named holders or permitted transferees, including 6.202M private-warrant shares. It is not proof of resale.
Public-warrant primary issuance	Up to 9.488M shares at $11.50	Potential issuer shares. The strike is far above $1.03, so this is not near-term supply without a large rally.
New senior note conversion	About 595,238 shares at $4.20 against $2.5M principal	Potential conversion shares, not current common shares. Accrued or contractual details could change the ultimate amount.
New common warrant	Up to 2.050M shares at $4.20	Potential shares, with resale registration promised. Exercise and resale are unverified.
Pre-existing FPA accounting	$11.005M share-subscription receivable and $8.104M derivative liability	The escrowed prepayment is not unconditionally available to the company. The number of shares that could be delivered under the FPA is not established by the August filings.
Shortfall-warrant amendment	Exercise price is the greater of the FPA termination price and $0.85	The amendment creates a price-sensitive claim. The current filing does not provide enough information to model its share quantity.

The August 3 FPA amendment resets the termination price each scheduled trading week to the lower of $12.00 and the prior week's Bloomberg VWAP. The separate shortfall-warrant amendment sets its exercise price at the greater of that termination price and $0.85. These terms make the financing stack more reflexive, but they do not establish that shares have been issued or sold.

Positioning score: 3/5. The structural supply evidence is strong. The external positioning evidence is incomplete. The short remains fail-closed until a locate, borrow, same-session liquidity, and exit plan are independently verified.

The Catalyst

The catalyst path is a sequence rather than one headline:

Next several sessions: Watch for a resale, exercise, or ownership filing that identifies actual shares sold, price, proceeds, and holder. A prospectus alone is not the event.
Registration chain: The new purchase agreement calls for an S-1 or S-3 for note-conversion and warrant shares. Its definitions include a 30-day filing deadline and a 60-day effectiveness deadline, extended to 120 days if the SEC comments. These are contractual timing windows, not evidence that the registration is already effective.
Weekly FPA reset: The weekly VWAP-based termination price and the $0.85 shortfall-warrant floor can change the economics of future settlement. The quantity and settlement state must be verified before assigning a share count.
Next cash and share-count filing: This is the adjudicating event. It should show whether the $2.05M purchase price reached cash, how many common shares are outstanding, whether new claims were registered or exercised, and whether operating burn improved.

The cleanest disconfirming sequence is therefore: actual resale or exercise, cash receipt and post-financing share count, then the next operating cash-flow disclosure. If the sequence instead shows material customer cash, no resale, and improving operating cash flow, the short thesis should be abandoned.

The Gap

The high-confidence observation is a capital-structure gap: the latest filed common denominator is 28.94 million shares, the registered secondary maximum is 26.55 million shares, and the company added a senior near-term note plus warrant claims. The high-confidence inference is that the price can remain fragile even if the AI operating narrative is genuine, because the claims compete for a small equity value and the FPA cash is not unconditionally available.

The lower-confidence part is the timing. The market may already be discounting the overhang. The August 4 22.56% decline could be the first repricing, short covering could reverse it, or neither event could produce actual supply. Without a confirmed holder sale and borrow data, “registered shares equal immediate float” would be false precision.

The Payoff Map

The targets below are analytical price scenarios, not promises or personalized financial advice. “Short return” is calculated as (reference price - target price) / reference price, using the $1.03 Nasdaq reference.

Scenario	Target price	Probability	Gross short return	What must be true
Top case for the short	$0.55	25%	+46.6%	Actual resale or financing pressure appears, no credible cash rescue arrives, and the next filing confirms a weak cash and share-count bridge.
Base case	$0.82	50%	+20.4%	The prospectus remains an overhang, no immediate rescue changes the denominator, and the stock digests the filing chain without a durable operating improvement.
Bottom case for the short	$1.75	25%	-69.9%	Cash arrives, a customer or strategic contract changes expectations, FPA settlement is favorable, or a borrow squeeze overwhelms the structural thesis.

Probabilities sum to 100%. The probability-weighted price return is approximately +4.4% before borrow, fees, slippage, taxes, and gap risk. That EV is fragile because the missing live market-structure variables are precisely the variables that determine whether a nominal short payoff can be captured.

Price Target and Probability Map

The short thesis has a positive price-only EV but not a positive executable EV. A borrow fee, locate failure, recall, halt, spread, market impact, and overnight gap can consume the modeled edge. The right interpretation is “research edge pending execution proof,” not “trade ready.”

The price invalidation level is $1.75, subject to a fundamental review. A print above it alone is a risk control, not proof that the company has repaired its cash conversion. The thesis should be abandoned if the price reclaims that level and the reclaim is supported by a reconciled cash receipt, durable commercial evidence, or a clean post-financing share count.

What Could Go Wrong

The stock has already fallen 22.56% and traded near its 52-week low. Shorting after a large decline can pay for the thesis while still losing to a reflexive bounce.
The $4.20 conversion and warrant prices are out of the money. Holders may prefer to wait, and the prospectus may never become a near-term selling event.
Fusemachines may receive the $2.05M purchase price, obtain additional equity-line or strategic funding, or settle the FPA without common-stock pressure. The 10-Q's $11.005M escrowed prepayment is conditional, not worthless.
A customer contract, AI deployment, or acquisition announcement can change the numerator faster than financing arithmetic can change the denominator.
A low-float squeeze, borrow recall, trading halt, or unavailable exit liquidity can turn a correct fundamental view into an unmanageable position.
The latest 28.94M share count predates the August financing. Adding new claims to it without a later filing would double count or misclassify instruments.

What Would Prove This Wrong

The short thesis is wrong, or not yet actionable, if any of the following occurs:

A post-August financing filing shows cash received, a reconciled common count, no meaningful resale, and operating cash flow that is improving rather than merely funded by new equity.
The FPA settles in cash or the company obtains non-dilutive capital that removes the near-term maturity pressure without creating equivalent common supply.
A material customer contract produces disclosed cash collection, revenue, or margin contribution rather than only an AI narrative.
FUSE reclaims $1.75 for three regular sessions on verified volume while the resale and share-count filings show no new selling pressure. This is a risk-control rule, not a claim about intrinsic value.
A responsible locate and borrow cannot be obtained at acceptable cost and recall terms. That does not falsify the capital-structure thesis, but it falsifies the trade expression.

Best Trade Strategy

Direction: Conditional short bias, research only.

Preferred instrument: FUSE common stock only after execution gates pass. A defined-risk put spread could limit gap risk, but no live options chain, spread, volume, or open-interest data was verified, so no options expression is recommended.

Common-stock stance: Watchlist. entry.price = null; execution.can_execute = false.

Entry reference: $1.03 dated August 4 Nasdaq close. A permissible research trigger would require a verified locate and borrow, a fresh regular-session quote, acceptable spread and depth, sufficient volume and exit liquidity, then failed continuation after an actual resale, exercise, or cash/share-count catalyst. Do not short the $0.9771 intraday low.

Targets and invalidation: Base target $0.82, extension target $0.55, price-risk control $1.75. Recalculate after every new filing because the current common denominator may have changed.

Timeline: The first 1 to 10 regular sessions are the highest-catalyst window for a resale, exercise, or financing-response move. The broader thesis runs through the next cash/share-count filing and the February 12, 2027 note maturity, but a short should not be held mechanically until maturity.

Do not trade if:

No locate, borrow availability, borrow cost, recall terms, or settlement capacity is verified.
The quote, spread, depth, venue status, volume, or exit liquidity is missing or abnormal.
The stock is making a fresh intraday low without a failed bounce.
A material customer contract, cash receipt, non-dilutive financing, or favorable FPA settlement appears.
The only evidence is the prospectus maximum, the latest down day, or a stale market-cap field.
Options are considered without a verified chain and open interest.

Monitoring checklist:

Check	Required evidence	Current state
Supply	Actual resale, exercise, or post-financing share count	Unverified
Cash	Receipt of the $2.05M purchase price and next cash balance	Unverified
Operations	Revenue collection, margin, and operating cash flow	Latest Q1 evidence is weak; next quarter not available
FPA	Settlement state, termination price, and share quantity	Partial terms disclosed; quantity and settlement unverified
Market structure	NBBO, depth, locate, borrow, short interest, options, dealer and fund flow	Unverified
Invalidation	$1.75 reclaim plus cash or commercial confirmation	Not triggered in this research pass

Bottom Line

FUSE is the best current U.S.-only short candidate because the fresh filings expose a specific disagreement between a small equity denominator, a large resale prospectus, a senior near-term note, and a weekly-reset financing mechanism. The same filings also show why the trade is dangerous: potential supply is not immediate float, the new fixed-price claims are out of the money, cash receipt is not yet confirmed in a later balance sheet, and live borrow and liquidity are missing.

Publish decision: Conditional Short Note. The research edge is plausible, the execution edge is unverified, and the correct current signal is no trade.

Research Quality Scorecard

Criterion	Score	Reason
Is the mispricing specific?	5/5	Registered potential supply, senior maturity, FPA reset, and cash conversion are separate and linked.
Is there evidence beyond narrative?	4/5	Fresh SEC filings and a dated Nasdaq tape support the thesis; post-financing confirmation is pending.
Is positioning supported or clearly labeled as uncertain?	3/5	Structural claims are sourced, but live borrow, short interest, options, and holder sales are missing.
Is there a catalyst or plausible closing mechanism?	5/5	Resale, exercise, registration, weekly reset, and next cash/share-count filings are observable.
Is the downside case described honestly?	5/5	Cash rescue, contracts, FPA settlement, recall, and squeeze paths are explicit.
Is the strongest counterargument included?	5/5	Out-of-the-money claims and the absence of actual resale are treated as central counterevidence.
Is the article useful even if the trade is not taken?	4/5	The denominator and monitoring sequence are reusable, but tradeability remains unresolved.
Are all factual claims sourced or marked as unverified?	4/5	Filing claims are sourced; market-structure and future cash claims are explicitly marked unavailable or unverified.
Total	35/40	Conditional Short Note, not an executable trade.

Sources

Fusemachines August 2, 2026 Form 8-K, financing and Meteora amendments (primary)
Fusemachines August 4, 2026 Form 424B3, secondary resale and incorporated March 31 Form 10-Q (primary)
Fusemachines Second Common Stock Purchase Warrant Amendment (primary)
Fusemachines Second Forward Purchase Agreement Confirmation Amendment (primary)
Nasdaq FUSE quote record and Nasdaq FUSE summary (market data, accessed 2026-08-05)
Cycurion July 31 warrant-inducement 8-K exhibit, Cycurion August 4 acquisition-close 8-K, and Nasdaq CYCU quote record (screened candidate evidence)
Nuwellis August 3 financing 8-K exhibit, Nuwellis August 3 Form 424B5, and Nasdaq NUWE quote record (screened candidate evidence)
60 Degrees Pharmaceuticals July 30 financing 8-K and Nasdaq SXTP quote record (screened candidate evidence)

AI Illustration Prompt

Create a restrained, high-end editorial financial illustration for The Mispricing Desk: a small glass tower labeled only with abstract geometric marks, surrounded by three transparent layers of paper representing filed common shares, registered resale claims, and a senior convertible note. A thin red thread connects a weekly VWAP dial to a $0.85 floor marker and a distant $4.20 conversion marker. Show a dark blue Nasdaq-like trading floor in the background with a sharp downward price line, then a countervailing upward spark to communicate squeeze risk. Use charcoal, ink blue, muted copper, and a single warning red accent. No real company logos, no readable ticker symbols, no numbers except the abstract markers, no candlestick clutter, no hype, no personalized financial advice. Include a subtle, tasteful The Mispricing Desk watermark in one corner.