Add DeepSeek-R1 model (#16631)

* Add DeepSeek-R1 model * Update Metadatas * Update CHANGELOG.md and optimise images --------- Co-authored-by: raycastbot <[email protected]>
raycast · Jan 30, 2025 · 3b0e033 · 3b0e033
1 parent a5cf0cd
commit 3b0e033
Show file tree

Hide file tree

Showing 11 changed files with 105 additions and 19 deletions.
diff --git a/extensions/deepseeker/CHANGELOG.md b/extensions/deepseeker/CHANGELOG.md
@@ -1,5 +1,11 @@
 # DeepSeeker - Quick Actions Changelog
 
+## [New Features] - 2025-01-30
+
+- 🐳 Add DeepSeek-R1 reasoning model
+- 🔄 Rename the deepseek-chat model to DeepSeek-V3 (official name)
+- 💸 Add token count and cost estimation for DeepSeek-R1
+
 ## [New Features] - 2025-01-19
 
 - 🙋 Add `Ask LLM` command using pure user query input.

diff --git a/extensions/deepseeker/README.md b/extensions/deepseeker/README.md
@@ -9,11 +9,11 @@ While [ChatGPT Quick Actions](https://www.raycast.com/alanzchen/chatgpt-quick-ac
 1. Get your DeepSeek API key from [DeepSeek API](https://platform.deepseek.com/api_keys). It's like `sk-37cd5***********************ac74`. Paste it in the `DEEPSEEK_API_KEY` environment variable in Raycast Settings.
 2. The `Custom API Endpoint` is optional. Only if you want to use a custom API endpoint such as ChatGPT's API. Leave it empty if you want to use the default DeepSeek API endpoint. A list of possible endpoints:
 
-    - DeepSeek: `https://api.deepseek.com/v1`
-    - ChatGPT: `https://api.openai.com/v1`
-    - Gemini: `https://generativelanguage.googleapis.com/v1beta/openai/`
+   - DeepSeek: `https://api.deepseek.com/v1`
+   - ChatGPT: `https://api.openai.com/v1`
+   - Gemini: `https://generativelanguage.googleapis.com/v1beta/openai/`
 
-3. Choose `deepseek-chat` as the global preferred model. You can also customize the preferred model for each command.(For now, DeepSeeker only has one model)
+3. Choose `DeepSeek-V3` or `DeepSeek-R1` (new reasoning model) as the global preferred model. You can also customize the preferred model for each command.
 
 Enjoy using DeepSeeker! 🚀
 
@@ -33,5 +33,15 @@ Enjoy using DeepSeeker! 🚀
   - [ ] Choose proopt when running the command
 - [ ] 💬 Conversation Chat mode
 
-
 > This extension is built highly inspired by [ChatGPT Quick Actions](https://www.raycast.com/alanzchen/chatgpt-quick-actions) by [Alan Chen](https://www.raycast.com/alanzchen).
+
+## Metadatas
+
+<details>
+<summary>Wallpaper & Examples</summary>
+
+The wallpaper for the Screenshots is the https://misc-assets.raycast.com/wallpapers/blue_distortion_2.heic
+
+All the examples have been generated on [Elon Musk's Wikipedia page 1st Paragraph](https://en.wikipedia.org/wiki/Elon_Musk)
+
+</details>
diff --git a/extensions/deepseeker/metadata/deepseeker-1.png b/extensions/deepseeker/metadata/deepseeker-1.png
diff --git a/extensions/deepseeker/metadata/deepseeker-2.png b/extensions/deepseeker/metadata/deepseeker-2.png
diff --git a/extensions/deepseeker/metadata/deepseeker-3.png b/extensions/deepseeker/metadata/deepseeker-3.png
diff --git a/extensions/deepseeker/metadata/deepseeker-4.png b/extensions/deepseeker/metadata/deepseeker-4.png
diff --git a/extensions/deepseeker/metadata/deepseeker-5.png b/extensions/deepseeker/metadata/deepseeker-5.png
diff --git a/extensions/deepseeker/package.json b/extensions/deepseeker/package.json
@@ -5,10 +5,19 @@
   "description": "Performs one-shot actions with DeepSeek and more",
   "icon": "icon.png",
   "author": "ViGeng",
+  "contributors": [
+    "degouville"
+  ],
   "categories": [
     "Productivity",
     "Developer Tools"
   ],
+  "keywords": [
+    "DeepSeek",
+    "DeepSeek V3",
+    "DeepSeek R1",
+    "LLM"
+  ],
   "license": "MIT",
   "commands": [
     {
@@ -38,9 +47,13 @@
               "value": "global"
             },
             {
-              "title": "deepseek-chat",
+              "title": "DeepSeek-V3",
               "value": "deepseek-chat"
             },
+            {
+              "title": "DeepSeek-R1",
+              "value": "deepseek-reasoner"
+            },
             {
               "title": "GPT-4o",
               "value": "gpt-4o"
@@ -92,9 +105,13 @@
               "value": "global"
             },
             {
-              "title": "deepseek-chat",
+              "title": "DeepSeek-V3",
               "value": "deepseek-chat"
             },
+            {
+              "title": "DeepSeek-R1",
+              "value": "deepseek-reasoner"
+            },
             {
               "title": "GPT-4o",
               "value": "gpt-4o"
@@ -146,9 +163,13 @@
               "value": "global"
             },
             {
-              "title": "deepseek-chat",
+              "title": "DeepSeek-V3",
               "value": "deepseek-chat"
             },
+            {
+              "title": "DeepSeek-R1",
+              "value": "deepseek-reasoner"
+            },
             {
               "title": "GPT-4o",
               "value": "gpt-4o"
@@ -225,9 +246,13 @@
               "value": "global"
             },
             {
-              "title": "deepseek-chat",
+              "title": "DeepSeek-V3",
               "value": "deepseek-chat"
             },
+            {
+              "title": "DeepSeek-R1",
+              "value": "deepseek-reasoner"
+            },
             {
               "title": "GPT-4o",
               "value": "gpt-4o"
@@ -279,9 +304,13 @@
               "value": "global"
             },
             {
-              "title": "deepseek-chat",
+              "title": "DeepSeek-V3",
               "value": "deepseek-chat"
             },
+            {
+              "title": "DeepSeek-R1",
+              "value": "deepseek-reasoner"
+            },
             {
               "title": "GPT-4o",
               "value": "gpt-4o"
@@ -333,9 +362,13 @@
               "value": "global"
             },
             {
-              "title": "deepseek-chat",
+              "title": "DeepSeek-V3",
               "value": "deepseek-chat"
             },
+            {
+              "title": "DeepSeek-R1",
+              "value": "deepseek-reasoner"
+            },
             {
               "title": "GPT-4o",
               "value": "gpt-4o"
@@ -383,7 +416,7 @@
           "default": "Answer the following question concisely: \n\n"
         },
         {
-          "name": "model_ask", 
+          "name": "model_ask",
           "description": "Override the global preferred model for this command",
           "type": "dropdown",
           "required": false,
@@ -395,9 +428,13 @@
               "value": "global"
             },
             {
-              "title": "deepseek-chat",
+              "title": "DeepSeek-V3",
               "value": "deepseek-chat"
             },
+            {
+              "title": "DeepSeek-R1",
+              "value": "deepseek-reasoner"
+            },
             {
               "title": "GPT-4o",
               "value": "gpt-4o"
@@ -411,7 +448,7 @@
               "value": "gpt-4"
             },
             {
-              "title": "GPT-4 Turbo", 
+              "title": "GPT-4 Turbo",
               "value": "gpt-4-turbo"
             },
             {
@@ -449,9 +486,13 @@
               "value": "global"
             },
             {
-              "title": "deepseek-chat",
+              "title": "DeepSeek-V3",
               "value": "deepseek-chat"
             },
+            {
+              "title": "DeepSeek-R1",
+              "value": "deepseek-reasoner"
+            },
             {
               "title": "GPT-4o",
               "value": "gpt-4o"
@@ -503,9 +544,13 @@
               "value": "global"
             },
             {
-              "title": "deepseek-chat",
+              "title": "DeepSeek-V3",
               "value": "deepseek-chat"
             },
+            {
+              "title": "DeepSeek-R1",
+              "value": "deepseek-reasoner"
+            },
             {
               "title": "GPT-4o",
               "value": "gpt-4o"
@@ -549,9 +594,13 @@
               "value": "global"
             },
             {
-              "title": "deepseek-chat",
+              "title": "DeepSeek-V3",
               "value": "deepseek-chat"
             },
+            {
+              "title": "DeepSeek-R1",
+              "value": "deepseek-reasoner"
+            },
             {
               "title": "GPT-4o",
               "value": "gpt-4o"
@@ -618,9 +667,13 @@
       "type": "dropdown",
       "data": [
         {
-          "title": "deepseek-chat",
+          "title": "DeepSeek-V3",
           "value": "deepseek-chat"
         },
+        {
+          "title": "DeepSeek-R1",
+          "value": "deepseek-reasoner"
+        },
         {
           "title": "GPT-4o",
           "value": "gpt-4o"

diff --git a/extensions/deepseeker/src/common.tsx b/extensions/deepseeker/src/common.tsx
@@ -103,6 +103,13 @@ export default function ResultView(prompt: string, model_override: string, toast
     getResult();
   }
 
+  async function retryWithDeepSeekReasoner() {
+    setModel("deepseek-reasoner");
+    setLoading(true);
+    setResponse("");
+    getResult();
+  }
+
   useEffect(() => {
     getResult();
   }, []);
@@ -154,6 +161,14 @@ export default function ResultView(prompt: string, model_override: string, toast
                 icon={Icon.ArrowNe}
               />
             )}
+            {model != "deepseek-reasoner" && (
+              <Action
+                title="Retry with DeepSeek Reasoner"
+                onAction={retryWithDeepSeekReasoner}
+                shortcut={{ modifiers: ["cmd", "shift"], key: "d" }}
+                icon={Icon.ArrowNe}
+              />
+            )}
             {sidenote}
           </ActionPanel>
         )

diff --git a/extensions/deepseeker/src/util.ts b/extensions/deepseeker/src/util.ts
@@ -44,6 +44,8 @@ export function estimatePrice(prompt_token: number, output_token: number, model:
     price = (prompt_token * 0.15 + output_token * 0.6) / 10000;
   } else if (model == "gpt-4o") {
     price = (prompt_token * 5.0 + output_token * 15.0) / 10000;
+  } else if (model == "deepseek-reasoner") {
+    price = (prompt_token * 2.0 + output_token * 2.5) / 10000;
   } else if (model == "deepseek-chat") {
     price = (prompt_token * input_price + output_token * output_price) / 10000;
     // * there is a tmeporary discount for deepseek-chat, we ignore it for now

diff --git a/extensions/deepseeker/tsconfig.json b/extensions/deepseeker/tsconfig.json
@@ -20,4 +20,4 @@
     "jsx": "react-jsx",
     "resolveJsonModule": true
   }
-}
+}
-Original file line number
+Diff line change
@@ Expand Up / @@ -20,4 +20,4 @@ @@
         "jsx": "react-jsx",
         "resolveJsonModule": true
       }
-    }
+    }