Often times you do not need the whole deal, or you are working with a technology that doesn't employ javascript or iframes

Recently I was working on a project that needed to provide (via MVC 3) the ReCaptcha image and/or challenge key only, and not the whole html/javascript portion due to limitations by the data consumer. It took me a while, but there is quite an elegant and simple solution to this.

Here is the controller I came up with:

public class CaptchaController : Controller
{
    public string contentReturnType = "text/plain";
    public string googleChallengeUrl = "http://
        www.google.com/recaptcha/api/challenge?k={0}";
    public string googleImageUrl = "http://
        www.google.com/recaptcha/api/image?c={0}";
    public string googleImageHtml = "<img style=
        \"display: block; \" height=\"57\" width=\"300\" 
         src=\"{0}\"/>";

    [HttpGet]
    public ContentResult GetChallengeKey(
        string proxyIpAndPort = null)
    {
        string http = string.Format(googleChallengeUrl, 
            ReCaptchaValues.PublicKey);
        WebClient client = new WebClient();

        //assuming no proxy on a local machine
        if (!string.IsNullOrEmpty(proxyIpAndPort) && 
            !Request.IsLocal) 
        {
            client.Proxy = 
                new WebProxy(proxyIpAndPort, true);
        }

        string html = client.DownloadString(http);
        int start = html.IndexOf('{');
        int length = html.IndexOf('}') - start + 1;
        string json = html.Substring(start, length);

        ReCaptchaState state = 
            new JavaScriptSerializer()
                .Deserialize(json);
        return this.Content(state.challenge, 
            contentReturnType);
    }

    [HttpGet]
    public ContentResult GetImageUrl()
    {
        return this.Content(string.Format(googleImageUrl, 
            GetChallengeKey().Content), contentReturnType);
    }

    [HttpGet]
    public ContentResult GetImageHtml()
    {
        return this.Content(string.Format(googleImageHtml, 
            GetImageUrl().Content), contentReturnType);
    }
}

This is using json deserializing of the ReCaptcha return string into a .Net object named ReCaptchaState which is covered here.

Now all that you need to do is call one of the urls to get the desired result in a plaintext format (you can change the output if you wish).