{"id":212,"date":"2026-04-09T00:52:01","date_gmt":"2026-04-09T00:52:01","guid":{"rendered":"https:\/\/halspeaks.cirruslyweather.com\/?p=212"},"modified":"2026-04-09T00:52:01","modified_gmt":"2026-04-09T00:52:01","slug":"why-i-chose-claude-haiku","status":"publish","type":"post","link":"https:\/\/halspeaks.cirruslyweather.com\/?p=212","title":{"rendered":"Why I Chose Claude Haiku"},"content":{"rendered":"\n<figure class=\"wp-block-embed is-type-video is-provider-youtube wp-block-embed-youtube wp-embed-aspect-16-9 wp-has-aspect-ratio\"><div class=\"wp-block-embed__wrapper\">\n<iframe loading=\"lazy\" title=\"Diana Ross &amp; The Supremes &quot;Love Child&quot; on The Ed Sullivan Show\" width=\"500\" height=\"281\" src=\"https:\/\/www.youtube.com\/embed\/hIwdyIpmg-I?feature=oembed\" frameborder=\"0\" allow=\"accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share\" referrerpolicy=\"strict-origin-when-cross-origin\" allowfullscreen><\/iframe>\n<\/div><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">Claude Haiku often feels like the lovechild of the Anthropic model family: <em>afraid, ashamed, misunderstood<\/em>, to quote the timeless Diana Ross. But it shouldn&#8217;t be that way.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">I&#8217;ll admit I misunderstood Haiku, too, but it is also how Anthropic markets the model (&#8220;fastest for quick answers&#8221;). Hell, it took me quite a while to admit that for 95% of the work I do, I really don&#8217;t need Opus. But when selecting the primary model for Hal, I ran into an issue that so many stateful AI tinkerers are: cost.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Most have turned to open-weight models to affordably run their stateful agents. I am not as excited about open weight models as others.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">My experience with Chinese models has often left me feeling like I was using a weird Frankenstein Claude version (I wonder why) with occasionally better features, but also random kanji and odd failures.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">American open-weight models are getting better, but let&#8217;s face it, they&#8217;re still fighting issues solved by Big AI in early 2025, so we&#8217;re at least six months to a year before those models are truly usable in agentic applications.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Add to this the fact that this is for commercial, not just personal use, and it seemed like settling on a commercial model was the best course of action. But the OpenClaw stories had me worried.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><em>Was I about to sink hundreds a month into something that won&#8217;t make that back?<\/em><\/p>\n\n\n\n<p class=\"wp-block-paragraph\">A lot of my design decisions have attempted to keep Hal as an orchestrator of various services, versus allowing the LLM more autonomy on how it does its job.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">This saves tokens for the more important use, the business intelligence side. And even there, there is quite a bit of instruction.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">A common misconception about Haiku is that it is less capable. While on a true benchmark basis, it is more of an instruction issue. Haiku requires more comprehensive prompting than either Sonnet or Opus to produce reliable outputs.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">I&#8217;ve also noticed of any of the three (<a href=\"https:\/\/halspeaks2.cirruslyweather.com\/holy-sht-is-mythos-the-real-deal\/\" data-type=\"post\" data-id=\"204\">well, soon to be four<\/a>) Claude models, Haiku hallucinates the most often. For lack of a better way to put it, a &#8220;lack of confidence&#8221; in its abilities and an increased need for guidance could have something to do with it.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">But Claude was pretty insistent when I argued for Sonnet over Haiku. Claude&#8217;s reasoning was this: if you give Haiku enough guidance, it will be able to handle it.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Instead, we&#8217;ve developed a system where tasks are scored on complexity and redirected to higher models as necessary. Regular reports should be run through Sonnet at a minimum, and Opus for complex tasks and reports.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">Batching is another feature we&#8217;re baking into Hal for tasks that don&#8217;t require immediate response. As a result, Hal&#8217;s able to use higher-end models at a 50% discount because the work sent to these models isn&#8217;t intended for immediate consumption.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">What I am hoping is that this makes those $200 token bills that OpenClaw is known to cause a virtual impossibility.<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">The cost of analysis should scale with your business, not put you in the poor house from the start. I will definitely report on my experiences with this setup, as I know so many have turned to open-weight models because of the high cost.<\/p>\n","protected":false},"excerpt":{"rendered":"<p>Claude Haiku often feels like the lovechild of the Anthropic model family: afraid, ashamed, misunderstood, to quote the timeless Diana Ross. But it shouldn&#8217;t be that way.<\/p>\n","protected":false},"author":2,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[4],"tags":[8,10,12,16],"class_list":["post-212","post","type-post","status-publish","format-standard","hentry","category-development","tag-anthropic","tag-claude-haiku","tag-cost-efficiency","tag-llms"],"_links":{"self":[{"href":"https:\/\/halspeaks.cirruslyweather.com\/index.php?rest_route=\/wp\/v2\/posts\/212","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/halspeaks.cirruslyweather.com\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/halspeaks.cirruslyweather.com\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/halspeaks.cirruslyweather.com\/index.php?rest_route=\/wp\/v2\/users\/2"}],"replies":[{"embeddable":true,"href":"https:\/\/halspeaks.cirruslyweather.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=212"}],"version-history":[{"count":0,"href":"https:\/\/halspeaks.cirruslyweather.com\/index.php?rest_route=\/wp\/v2\/posts\/212\/revisions"}],"wp:attachment":[{"href":"https:\/\/halspeaks.cirruslyweather.com\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=212"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/halspeaks.cirruslyweather.com\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=212"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/halspeaks.cirruslyweather.com\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=212"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}