Let's begin. We'll start with a circle of any given radius $r$. (We could do the proof just in the special case $r=1$, but for completeness we'll keep it arbitrary).
Figure 1: Arbitrary circle
When Archimedes began to approximate $\pi$ thousands of years ago, he took as a given (we aren't) that $\pi$ was the area of the unit circle and aimed to approximate that instead. To do this, he inscribed a regular polygon with some number $N$ of vertices into the unit circle such that each vertex lay on the circle itself. He would then calculate the area of this polygon. As he increased $N$, he would get better and better lower bounds on $\pi$ but, with Greek methods, the area became increasingly hard to compute.
Figure 2: Inscribed 8-gon
Above is the inscription of an 8-gon ($N=8$). Our goal is to use trigonometry to find the total area of this inscribed polygon in terms of $N$, and then take the limit as $N\to\infty$. Our method will be to divide the $N$-gon into $N$ triangles by drawing radii from the center-point $A$ to each vertex, and then multiplying by $N$ to get the total area as shown.
Figure 3: Dividing into triangles showing angle $\theta$
To find the area of one triangle, we will use the famous trigonometric area formula $$A = \frac 1 2ab\sin\theta,$$ where $a$ and $b$ are two sides of the triangle and $\theta$ is the angle between them. (Deriving this is a simple exericse in trigonometry: recall that $A=\frac 1 2bh$ and express $b,h$ in terms of $\sin\theta$ and two lengths). In our case, we can see that $a=b=r$ (the radius of the circle). What about $\theta$? If we look at all the triangles in question, we note that the sum of $N$ total copies of $\theta$ should give us one full rotation of $2\pi$. Hence we can simply divide by $N$ to get the value of $\theta=\frac{2\pi}N$. Plugging this all in and multiplying by $N$, we find the area of the inscribed $N$-gon to be $$A_N = N\cdot\frac 1 2ab\sin\theta = \frac N 2r^2\sin\left(\frac{2\pi}N\right).$$ Thus, when we send $N\to\infty$ we should get better and better approximations to the area and the true area should come in the limit. Thus we need to compute $$\lim_{N\to\infty}A_N = \lim_{N\to\infty}\frac{Nr^2}2\sin\left(\frac{2\pi}N\right) = \frac{r^2}2\lim_{N\to\infty}N\sin\left(\frac{2\pi}N\right).$$ To evaluate the rightmost limit, set $n=\frac{2\pi}N$. Then $N=\frac{2\pi}n$, and as $N\to\infty$ the multiplicative inverse $n\to 0.$ Plugging this in, $$A = \frac{r^2}2\lim_{n\to 0}\frac{2\pi}n\sin n = \pi r^2\lim_{n\to 0}\frac{\sin n}n.$$ But, lo and behold, this is our result from above - this final limit is $1$! Thus, we see as desired that in the limit $$A = \pi r^2$$ and we are complete!
In conclusion, this is one of the most beautiful proofs of the circle area formula, if you ignore the bit of limit algebra at the end. The main idea of the proof, inscription of regular polygons, dates back thousands of years to the Greeks. All we have done here is apply some more modern methods to this classical insight to get a general formula, which truly shows just how powerful even the most ancient mathematical methods can be.