{"id":28323,"date":"2025-02-08T10:48:12","date_gmt":"2025-02-08T02:48:12","guid":{"rendered":"https:\/\/www.1ai.net\/?p=28323"},"modified":"2025-02-08T10:48:12","modified_gmt":"2025-02-08T02:48:12","slug":"%e6%89%a3%e5%ad%90%e4%b8%8a%e7%ba%bf%e6%bb%a1%e8%a1%80%e7%89%88deepseek%e5%85%a8%e5%ae%b6%e6%a1%b6%ef%bc%8c%e7%94%a8deepseek%e6%99%ba%e8%83%bd%e4%bd%93%e6%89%93%e9%80%a0%e7%88%86%e6%ac%be%e7%9f%ad","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/28323.html","title":{"rendered":"Buckle goes live with full-blooded version of Deepseek family bucket to create explosive short videos with DeepSeek smart body"},"content":{"rendered":"

Just now.button<\/a>It's official, Buckle is available for all users and has support for theDeepSeek<\/a> R1\/V3 models are used for a free experience.<\/p>\n

DeepSeek-R1 model<\/strong><\/p>\n

The model thinks deeply through the chain of thought before generating the final response.<\/p>\n

Free to use, limited to 20 conversations\/day per user. For more conversations, you can turn on Buckle Pro and create a model inference access point for DeepSeek-R1, billed per Token usage.<\/p>\n

\"c46a9392p00srcfmn001bd000in0030m\"<\/p>\n

DeepSeek-V3 model<\/strong><\/p>\n

The model has enhanced language comprehension and generation capabilities.<\/p>\n

Free to use, limited to a maximum of 50 conversations\/day per user.<\/p>\n

\"6ed2f886p00srcfmn0019d000ig002tm\"<\/p>\n

DeepSeek-R1-Distill-Qwen-7B<\/strong><\/p>\n

Built through the inference distillation technique of the DeepSeek-R1 660B model. Excels in complex tasks such as mathematical reasoning and code generation, with particular expertise in multi-step derivation and cross-domain problem solving.<\/p>\n

Free to use, limited to a maximum of 100 conversations\/day per user.<\/p>\n

\"4dd4af89p00srcfmn000od000bg001qm\"<\/p>\n

DeepSeek-R1-Distill-Qwen-32B<\/strong><\/p>\n

Built through the inference distillation technique of the DeepSeek-R1 660B model; it excels in complex tasks such as mathematical inference and code generation, and is especially good at multi-step derivation and cross-domain problem solving, with several metrics exceeding OpenAI o1-mini.<\/p>\n

Free to use, limited to a maximum of 80 conversations\/day per user.<\/p>\n

\"0c0b0903p00srcfmn000pd000bf001vm\"<\/p>\n

If you don't know how to useAgent<\/a>If you just want to experience DeepSeek, you can follow this method and simply configure it.<\/p>\n

go intoBuckle official website<\/a>, add a new intelligent body.<\/p>\n

\"6ae74a3cj00srcfmp00agd000u000bvm\"<\/p>\n

Creating intelligences is fine.<\/p>\n

\"c17edb47j00srcfmp00bjd000u000ibm\"<\/p>\n

Just name it whatever you want.<\/p>\n

\"94a0cb77j00srcfmn0010d000db00c5m\"<\/p>\n

Then the model selection DeepSeek R1 or any other DeepSeek model will do.<\/p>\n

\"347a0393j00srcfmo004cd000fs00irm\"<\/p>\n

Nothing else needs to be configured to ask questions in the right panel, and as you can see from the debugging details on the right, the DeepSeek R1 model is indeed used.<\/p>\n

\"aeeb373ej00srcfmp009wd000u000j9m\"<\/p>\n

The above is a basic instruction, mainly for some of you who have never used buckles.<\/p>\n

Well, let's try to make a smart body with Button Coze + DeepSeek that can generate pop-ups in one click!Short Video<\/a>Text.<\/p>\n

initial step<\/strong><\/p>\n

Create a new workflow, then after the start node, insert a large model node, which can be named \"Link Extraction\". The purpose is to extract the short video links output by users.<\/p>\n

\n
    \n
  • Hint word: Extract the links in the string, keeping only the link part.<\/li>\n<\/ul>\n<\/blockquote>\n

    \"54f537c7j00srcfmp005hd000u000i6m\"<\/p>\n

    second step<\/strong><\/p>\n

    After the link extraction node, insert a plugin node, this plugin is called: link extraction<\/p>\n

    Link: https:\/\/www.coze.cn\/store\/plugin\/7329410795979161663<\/p>\n

    This plugin is a web content crawler. It is able to return the raw content of a web page, including title, content, links, etc., and it can also filter and parse this content.<\/p>\n

    \"b2b9ed51j00srcfmo0048d000ng00fbm\"<\/p>\n

    third step<\/strong><\/p>\n

    The content extracted back by the plugin may have some useless content, so you can add a big model node to accurately extract the original copy out.<\/p>\n

    Tips:<\/p>\n

    # Role
    \nYou're an accurate text analyst, able to quickly analyze text and extract the copy in it while keeping the original text intact.<\/p>\n

    ## Skills
    \n### Skill 1: Accurately identify content
    \nAccurately return the corresponding headlines, and copy.<\/p>\n

    The content of [title] in the content corresponds to the title
    \nThe content of [Video ASR Text] in the content corresponds to the main text<\/p>\n

    Output format:
    \nTitle: xxxx
    \nText: xxxx<\/p>\n

    ### Skill 2: Body Content Recognition
    \nThe body of the text eliminates elements such as title, time of publication, author, etc.<\/p>\n

    To save on the number of times DeepSeek is used, the beanbag model can be used here.<\/p>\n

    \"7be701acj00srcfmp005cd000ny00mbm\"<\/p>\n

    fourth step<\/strong><\/p>\n

    Rewrite the new pop-up copy with DeepSeek, insert a new node, and rewrite the original copy with DeepSeek.<\/p>\n

    For example, if I'm going to make short oral videos, I'll use DeepSeek short video-specific cue words.<\/p>\n

    Please rewrite a new copy based on the short video copy provided according to the pop-up structure, modularized and the copy must contain four modules.<\/p>\n

    Module 1: Hook Starters, Capturing the Audience's Attention with a Sentence
    \nThe second module: pain point resonance, so that the user instantly generates a sense of empathy
    \nThe third module: to solutions, must give solutions, to the audience to feel that you are reliable
    \nModule 4: To lead your audience to give you some attention and even pay you for it<\/p>\n

    Eventually, please just output the complete new copywriting<\/p>\n

    \"0dafbda3j00srcfmp005ed000o900mmm\"<\/p>\n

    Just output the output at the end.<\/p>\n

    This is the complete workflow:<\/p>\n

    \"922ddd6fj00srcfmo0043d000u000bkm\"<\/p>\n

    Let's take a final look at the results, above is a link to a short video I found on Shakeology.<\/p>\n

    \"8f8b44e3j00srcfmo005ad000ic00frm\"<\/p>\n

    The intelligent body then directly extracts the original text with a single click and rewrites a new pop-up copy.<\/p>\n

    \"0fd90dc3j00srcfmo0037d000hn00emm\"<\/p>\n

    Doing traffic with DeepSeek has really gotten easier.<\/p>\n

    Did you learn?<\/p>","protected":false},"excerpt":{"rendered":"

    Just recently, it was officially announced that buttons were open to all users and that they had been used to support the DeepSeek R1\/V3 model, which was free of charge. The DeepSeek-R1 model allows for in-depth thinking through the chain of thought before generating a final response. Free of charge, with a maximum of 20 conversations\/days per user. If more conversations are needed, it is possible to open a professional button version and create a model DeepSeek-R1 model reasoned access point, based on Token usage. DeepSeek-V3 Model Models have greater language understanding and generation. Free use, with a maximum limit of 50 conversations\/days per user. DeepSeek-R1-Distil-Qwen-7B<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"closed","ping_status":"closed","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[149,144],"tags":[3606,2445,1355,1007],"collection":[5669],"class_list":["post-28323","post","type-post","status-publish","format-standard","hentry","category-jiaocheng","category-baike","tag-deepseek","tag-2445","tag-1355","tag-1007","collection-deepseek"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/28323","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=28323"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/28323\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=28323"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=28323"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=28323"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=28323"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}