Monday, September 14, 2026
MiningNewsTerminal
Monday, September 14, 2026 Admin

BRAZ.CN ·

Canary Gold Corp. Reports Visible Gold Grains in Gravity Concentrates from 29 of 30 One-Metre RC Samples at Madeira River Project

CANARY GOLD REPORTS VISIBLE GOLD GRAINS IN GRAVITY

CONCENTRATES FROM 29 OF 30 ONE-METRE RC SAMPLES AT

MADEIRA RIVER PROJECT

Independent ODM results support the transition to representative whole-sample testwork and focused

sonic sampling at the Jaci Paraná target.

Vancouver, British Columbia – August 14, 2026 – Can ary Gold Corp. (CSE: BRAZ; OTCQB: CNYGF;

Frankfurt: K5D) (“Canary Gold” or the “Company”) is pleased to report visible-gold-grain observations

from 30 consecutive one-metre reverse-circulation ( “RC”) samples collected from RC-0032 at the Jaci

Paraná target within the Company’s Madeira River Pr oject in Rondônia, Brazil. Overburden Drilling

Management Limited (“ODM”) identified visible gold grains in laboratory-prepared gravity

concentrates from 29 of the 30 samples. The results demonstrate repeated downhole recovery of visible

gold grains at RC-0032, but do not represent head g rades or establish grade continuity beyond the

sampled hole.

Highlights

• Independent specialist laboratory Overburden Drilling Management Limited (“ODM”) identified visible

gold grains in laboratory-prepared gravity concentr ates from 29 of 30 consecutive one-metre RC

samples collected from RC-0032.

• ODM identified 190 visible gold grains. Of these , 156 grains (82.1%) were reshaped, consistent with

transport, abrasion and alluvial reworking within an auriferous sedimentary system.

• The visible-gold population is fine by particle count: 166 grains (87.4%) measured 100 micrometres

(µm) or less. However, only five grains larger than 150 µm accounted for approximately 44.1% of the

calculated visible-gold mass reported for the conce ntrate, illustrating a pronounced coarse-particle

tail and material nugget-effect risk.

• The ODM observations are geologically meaningful but semi-quantitative. Reported calculated ppb

values relate to visible gold in a heavy-mineral co ncentrate and are based partly on an assumed

concentrate mass. They are not sample head grades, metallurgical recovery results or mineral

resource estimates.

• The RC-0032 results provide additional vertical information at one location within the approximately

6,700-hectare conceptual paleochannel target. Resul ts received from the broader exploration

program support the paleochannel interpretation but do not yet establish representative grade or

grade continuity.

• Canary is planning an intermediate quantitative test program comprising approximately 40 selected

one-metre RC samples expected to weigh approximately 25 kilograms each. The proposed work would

include controlled gravity concentration, analysis of the gravity product and tails, and mass-balance,

volume and density controls, subject to final laboratory scope and Qualified Person review.

• The Company received an indication of potential sonic-drill availability in September 2026 and is

evaluating contracting, site access, permitting, pr ogram design and budget requirements. Focused

sonic sampling would be designed to assess represen tative recovered volumetric gold grades,

expressed in milligrams per cubic metre (mg/m³), while preserving both fine gold and the important

coarser-gold fraction.

• A larger sonic drilling campaign, potentially ex ceeding 15,000 metres, would only be considered if

representative sampling produces sufficiently encou raging and reproducible grade, recovery,

continuity, density and volume data to justify resource delineation work.

Figure 1. Illustrative visible gold grains recovered by ODM from laboratory-prepared gravity

concentrates from Batch 4028. A. Sample 3075: RC-0032 (27m – 28m), B. Sample 3076: RC-0032 (28m

– 29m). The 100 µm scale bars illustrate the fine particle dimensions and variable grain morphology.

Source: Overburden Drilling Management Limited. Photographs show visible gold grains recovered from

the submitted material; they do not demonstrate sample head grade, metallurgical recovery or economic

grade.

A Large Prospective Paleochannel Moving from Discovery to Evaluation

Canary’s 2026 exploration program has established a substantial body of geological evidence supporting

a broad paleochannel interpretation at the Jaci Paraná target within the Madeira River Project. The area

currently encompassed by the Company’s conceptual in terpretation is approximately 6,700 hectares,

beneath approximately 15 to 25 metres of overburden , with prospective alluvial sedimentary packages

locally interpreted to reach approximately 10 to 30 metres in thickness. These dimensions remain

conceptual and are not a mineral resource or volume estimate.

The Company’s exploration emphasis is transitioning from discovery—demonstrating the presence and

distribution of gold within sampled paleochannel material—toward technical evaluation. The immediate

objective is to determine whether representative sa mpling can produce reproducible volumetric-grade

and recovery information sufficient to justify further evaluation.

“The significance of the ODM results is the repeate d recovery of visible gold grains through the submitted interval in RC-

0032, rather than any single calculated concentrate va lue,” said Mark Tommasi, President of Canary Gold. “The results

provide additional vertical information at that loc ation and support continued evaluation, but they ar e not representative

head grades or metallurgical recovery results and d o not establish grade continuity beyond the sampled hole. Our next

phase is designed to generate representative and au ditable volumetric data. We plan to test complete s ubmitted RC

sample material, quantify gravity recovery and losses and then evaluate focused sonic sampling. A majo r resource

delineation program would only be considered if tha t work produces sufficiently encouraging and reprod ucible results.”

— Mark Tommasi, President

Figure 2. Correlated conceptual geological model for the Jaci Paraná target. The interpretation shows

a north-south paleochannel axis, correlated alluvial fill and gravel and coarse-sand lenses interpreted

to have relative continuity above irregular granitic basement.

Source: Company interpretation based on geological sections and drilling control. Preliminary conceptual

model only; do not use for resource or volume estimation.

Independent ODM Results Demonstrate Repeated Visible-Gold-Grain Recovery at RC-0032

ODM Batch 4028 comprised 30 consecutive one-metre R C samples, numbered 3075 to 3104, collected

from RC-0032 and consisting of sand and sand-plus-g ravel material. ODM processed the samples and

reported process weights, visible-gold-grain counts, morphology and particle dimensions. The

observations independently demonstrate repeated vis ible-gold-grain recovery from the submitted

interval at RC-0032; they do not independently establish the lateral extent of the paleochannel or grade

continuity beyond that hole.

Visible gold grains were recovered from gravity concentrates prepared from 29 of the 30 samples, with a

total of 190 grains identified. Samples 3075 to 3080 contained 101 of the 190 grains, while samples 3086

and 3087 form a second interval of interest. Sample 3087 contained six of the ten pristine grains identified

in the batch, potentially indicating local variatio n in sediment source or transport history. Sample 3 084

(36 m–37 m) was the only sample without identified visible gold.

Table 1. ODM Batch 4028 - Results at a Glance

Indicator Result Technical interpretation

Analysis scope 30 complete 1 m RC samples Full ODM Batch 4028 dataset reviewed.

Positive samples 29/30 (96.7%) Visible gold grains recovered from 29 submitted samples; not proof of

economic grade or continuity.

Visible grains 190 A geological observation, but not a total-gold assay or metallurgical

recovery result.

Morphology 156 reshaped (82.1%) Dominant signature of transport, abrasion and alluvial reworking.

Other morphologies 24 modified; 10 pristine Sample 3087 contains six of the ten pristine grains.

Median maximum length 50 µm Fine particles dominate the population by count.

Length ≤100 µm 166 grains (87.4%) Sampling and processing must preserve the fine fraction.

Length >150 µm 5 grains (2.6%) Rare coarse grains create material sampling and nugget -effect risk.

Calculated coarse mass share 44.1%* Five coarse grains contribute nearly half of calculated visible-gold mass

reported for the concentrate.

Maximum reported grain ~250 × 400 × 50 µm One grain in sample 3087 accounts for approximately 94% of its calculated

visible-gold value.

*Calculated visible-gold mass uses the reported visible-gold value in heavy-mineral concentrate and an assumed nonmagnetic concentrate

weight. It is not sample head grade or a metallurgical recovery result.

Particle Size, Recovery Risk and the Nugget Effect

The ODM particle-size distribution explains why the present results are important geologically but cannot

yet be used economically. By count, the population is dominated by fine gold: 87.4% of grains are 100 µm

or less and the median maximum grain length is 50 µm. This fine fraction may be vulnerable to loss during

RC drilling, sample handling and gravity concentration if recovery conditions are not specifically designed

to preserve it.

At the same time, the calculated visible-gold mass reported for the concentrate is strongly influenced by

a very small number of larger grains. Only five grains above 150 µm represented approximately 44.1% of

that calculated visible-gold mass. In sample 3087, one grain measuring approximately 250 × 400 × 50 µm

accounts for approximately 94% of the sample’s calcu lated visible-gold response; in sample 3078, two

coarser grains account for approximately 71%. Grain count is therefore not proportional to the calculated

visible-gold mass reported for the concentrate.

This combination - abundant fine particles and a ra re coarse-particle tail - creates two distinct technical

requirements. Sampling and processing must preserve fine and laminar gold while also using enough

sample support to capture the low-frequency coarse particles that can materially influence average

recovered grade. This is the central reason Canary is moving toward complete-sample testwork and

representative-volume sonic drilling.

CRITICAL TECHNICAL BOUNDARY

ODM measured process weights and visible gold grains and then calculated visible-gold values in a

heavy-mineral concentrate using an assumed nonmagnetic concentrate mass equivalent to 0.4% of

table feed. The reported ppb values refer to calculated visible gold in that concentrate. They are not

total gold in the full RC sample, are not head grades, do not establish metallurgical recovery and

should not be used to estimate mineral resources or economic value.

Widespread Gold Occurrence Supports the Paleochannel Model

The current exploration database includes 45 RC holes totalling 2,851 metres, 14 PQ (dry diamond drilling)

holes totalling approximately 902 metres and 63 screw auger holes totalling approximately 1,297 metres,

for a combined 2026 drilling and auger program of a pproximately 5,050 metres. Drilling has intersected

repeated intervals of rounded to subrounded granule s, pebbles, gravels and coarse sands across the

target area, consistent with the interpreted alluvial architecture.

Gold particles have been observed in concentrated material from multiple drill locations. The ODM Batch

4028 results independently demonstrate repeated vis ible-gold-grain recovery through the submitted

interval from RC-0032 and provide additional vertical information at that location.

The broad spatial distribution of drilling provides geological context across a target measured in

kilometres. However, individual RC holes sample a v ery small volume relative to the conceptual

paleochannel target, and the ODM batch discussed in this release comes from a single hole. The Company

is therefore using the existing drilling to prioritize representative sampling areas rather than to infer grade

continuity between widely spaced holes.

Figure 3. Current drill coverage across the interpreted Jaci Paraná paleochannel and Canary tenure.

Completed drill holes are shown across a broad target area and provide the basis for selecting focused

validation and sonic sampling locations.

A Staged Program to Establish Representative Volumetric Grade

Canary’s next work program is deliberately staged to test representativeness and recovery before the

Company commits capital to a major resource delineation campaign. Each phase is intended to answer a

specific technical question and to provide a clear decision gate for the next level of expenditure.

1 Whole -sample RC validation testwork

Approximately 40 selected one-metre RC samples, eac h expected to weigh

approximately 25 kilograms, are proposed for quanti tative metallurgical screening,

subject to confirmation of laboratory scope and Qua lified Person review. The

proposed program would use controlled gravity concentration followed by analysis

of the gravity product and tails. Wet and dry weights, feed mass, concentrate mass,

recovery by size fraction, density, volume and anal ytical mass balance would be

recorded. This phase is intended to provide a more quantitative estimate of gold

recovered from each complete submitted sample, assess gold remaining in the tails,

identify losses associated with the current discove ry-stage procedure and inform

selection of priority intervals for sonic sampling.

2 Focused sonic drilling and bulk sampling

The Company has received an indication of potential sonic-rig availability beginning

in September 2026 and is evaluating contractor, sit e, access, permitting, program-

design and budget requirements. The initial design under consideration comprises

close-spaced sonic grids across selected priority a reas, with each grid potentially

involving approximately 800 metres of drilling. Sonic drilling is expected to improve

interval recovery and sample-volume control while preserving sands, fines and rare

coarse particles. Multiple priority areas may be te sted, subject to the initial whole-

sample results, access, permitting, contracting, financing and budget approval.

3 Representative Recovered Gold per Unit Volume

Each sonic interval would be processed using a documented and auditable protocol

designed to assess recovered gold per controlled sa mple volume. The principal

reporting metric is expected to be milligrams of re covered gold per cubic metre

(mg/m³), supported by density, moisture, sample rec overy and mass-balance data.

Final terminology and conversion to an in-place volume basis will be reviewed by the

Qualified Person before implementation.

4 Economic decision gate before resource drilling

Only if representative sampling produces sufficiently encouraging and reproducible

grade, recovery, continuity, density, volume and mi ning-characteristic data would

Canary consider a larger sonic drilling campaign. S ubject to those results, the

Company anticipates that a resource delineation phase could exceed 15,000 metres

and would be designed to collect data required to a ssess whether an inferred

mineral resource estimate can be prepared in accordance with National Instrument

43-101.

Figure 4. Canary’s decision-gated technical pathway. Progression toward resource delineation is

conditional on representative grade, recovery, continuity, density and volume control, QA/QC and a

supportable economic case.

Near-Term Work and News Flow

Approximately 50 additional samples are currently b eing processed by ODM. Further SGS batches from

RC concentrate work are also pending or in progress. Canary intends to use these discovery-stage results

to refine the geological model, identify priority in tervals and improve sampling procedures; however,

concentrate observations and values will not be rep orted as equivalent to head grades or metallurgical

recovery results.

The proposed whole-sample testwork is expected to require approximately three to four weeks following

laboratory receipt, subject to confirmation of the f inal scope, contracting and laboratory scheduling.

Results would be used to inform the location and co nfiguration of initial sonic grids and to refine th e

recovery protocol before any field mobilization decision.

The Company will provide further updates as material results are received and the sonic sampling program

is finalized. Any decision to expand to resource delineation would be based on representative quantitative

data rather than visible-gold observations alone.

Quality Assurance, Quality Control and Analytical Methodology

ODM Batch 4028 comprised 30 consecutive one-metre RC samples of sand and gravel material from RC-

0032 that were submitted to Overburden Drilling Man agement Limited (“ODM”) at its laboratory in

Ottawa (Nepean), Ontario, Canada. ODM recorded bulk received weights, retained an approximately 300-

gram archival split, screened and processed sample material, and examined pan concentrates for visible

gold and fine-grained metallic indicator minerals. Visible gold grains were counted, classified by

morphology and measured. ODM’s calculated visible-g old values are based in part on an assumed

nonmagnetic heavy-mineral-concentrate mass equivalent to 0.4% of table feed and are therefore semi-

quantitative. The ODM work was not a metallurgical recovery test.

SGS work referenced in the Company’s technical repor ting relates to centrifuge or heavy-mineral

concentrates that reduce the original sample mass b y approximately 100-fold to 200-fold. Those

concentrate values do not represent the gold grade of the original bulk sample. The proposed validation

program is intended to improve quantitative charact erization by testing substantially larger complete

submitted samples, recording sample volume and density, applying mass-balance controls and analyzing

both recovered product and tails. Any SGS results used to support public geological conclusions should be

disclosed with applicable sample information, methods and QA/QC.

The Company is developing enhanced QA/QC protocols for the sonic program, including documented wet

and dry weights, interval recovery, volume, density, fe ed and product masses, duplicates, blanks,

standards where appropriate, chain of custody and recovery by particle-size fraction. Final procedures will

be reviewed by the Company’s Qualified Person before implementation.

Qualified Person and Technical Disclosure

The scientific and technical information contained in this news release has been reviewed and approved

by Andrew Lee Smith, P.Geo., a Qualified Person as defined by National Instrument 43-101 – Standards

of Disclosure for Mineral Projects.

Mr. Smith has previously visited Canary’s Madeira River project in Brazil and is familiar with the regional

geology and exploration activities. While not continuously present on site, Mr. Smith has reviewed drilling,

sampling, QA/QC and geological information and is satisfied that the information has been collected and

reported by site-based personnel in accordance with industry standard practices.

Mr. Smith is not independent of the Company within the meaning of National Instrument 43-101, as he

serves as Executive Chairman of Canary Gold Corp.

About Canary Gold Corp.

Canary Gold Corp. is a Canadian public exploration company focused on the acquisition and

advancement of gold projects in Brazil. The Company holds an option to earn up to a 70% undivided

interest in the Rio Madeira Project through a series of staged exploration expenditures and milestone

payments.

In August 2025, Canary further expanded its regional strategy by entering into a definitive agreement to

acquire a 100% interest in ten additional mineral t enements totaling approximately 94,700 hectares

from Talisman Venture Partners Ltd. , a private British Columbia corporation. The total consideration of

CAD $1.7 million has been satisfied through staged cash and share payments.

Talisman retains a 1.0% net smelter return (NSR) ro yalty on future production from the acquired

tenements, one-half of which (reducing the NSR to 0 .5%) may be repurchased by the Company at any

time for CAD $1.0 million.

Together, these interests provide Canary Gold with a dominant and strategically consolidated land

position in the Madeira River region of Rondônia St ate — one of Brazil’s most prospective yet

underexplored gold provinces.

For Further Information, Please Contact:

Canary Gold Corp.

Mark Tommasi, President

Tel: (604) 318-1448

www.canarygold.ca

Cautionary Statement Regarding Forward-Looking Information

This news release contains forward-looking informat ion within the meaning of applicable securities law s, including

statements regarding additional ODM and SGS work; t he proposed whole-sample test program and its scope ,

methods, timing and objectives; potential sonic-dri ll availability, contracting and mobilization; site access and

permitting; the design, timing and scale of possible sonic sampling and resource delineation programs; the ability of

future work to generate representative volumetric-grade, recovery, continuity, density and volume information; and

the possible preparation of a mineral resource esti mate. Forward-looking information is based on manag ement’s