XingChen: Xing4.0 29B A4B (free)

{"about":"### China Telecom AI 的 Xing4.0\n\nXing4.0-29B-A4B 是 China Telecom Artificial Intelligence Technology 的文字 MoE 模型，作為 TeleChat 的下一代開發。模型含 290 億參數，每個 token 啟用約 40 億參數。開發商將其定位於複雜工程任務與連續代理情境。\n\nXing4.0 的特色是完全使用 Ascend NPU 加速器與 MindSpore 訓練。架構結合 mHC、MLA 與 MTP，包含 40 層、64 個路由專家、每個 token 選取四個專家，以及一個共用專家。這些資料描述原始模型的結構與研究基礎，並非處理請求時使用的硬體。\n\n### 文字形式的工程工作\n\n開發商資料強調多步驟規劃、程式碼工作與連續解題。在文字對話中，可用於準備程式修改計畫、解釋相依關係，或根據傳入程式碼分析錯誤。先說明目標、輸入資料及限制，再要求提出行動順序與各步驟預期結果。\n\n程式設計任務請提供相關實作、資料格式與預期行為範例。實用結果可包含說明、局部修改及驗證範例。在自己的環境執行後，可將觀察回傳至對話，繼續修訂方案。這種提問保留答案與特定任務的關聯，不需要將自主存取專案的能力加諸模型。\n\n### 上下文與請求資料\n\n在 Ailoot，Xing4.0 接受文字並傳回文字。上下文視窗為 262,144 個 token，輸出上限為 235,929 個 token。共同上下文包含指示、原始資料、歷史與答案。可加入數個相關片段：需求、文件、程式碼及驗證結果。請給予清楚名稱，並指明所需結果，例如比較、演算法說明、更新後程式碼或解題計畫。\n\n開發商指出原始模型的上下文可擴展至 512K，但此連線大小為 262,144 個 token。準備大型請求時應以此為準，並保留答案空間。\n\n### 整合至應用程式\n\n提供 Messages、Responses 與 Chat Completions 格式，便於整合至熟悉的客戶端，並安排多則連續訊息。此連線的結果為生成文字。執行建議程式碼、讀取檔案及使用外部工具，需要應用程式具備相應邏輯。原始模型架構本身不會在一般文字請求內提供這些動作。","alias_of":null,"aliases":[],"author":"Xingchen AGI","canonical_model":"xingchen-agi/xing4.0-29b-a4b:free","ctx":262144,"formats":["chat","messages","responses"],"id":"xingchen-agi/xing4.0-29b-a4b:free","input":["text"],"kind":"text","max_out":235929,"methods":[{"cache_read":null,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":262144,"family":"anthropic","fast_multiplier":1,"flex_multiplier":1,"fmt":"messages","http":"POST","img_in":null,"img_out":null,"in":0,"input":["text"],"label":"Anthropic Messages","max_out":235929,"max_tool_calls":null,"out":0,"output":["text"],"path":"/v1/messages","request_model_id":"xingchen-agi/xing4.0-29b-a4b:free","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]},{"cache_read":null,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":262144,"family":"openai","fast_multiplier":1,"flex_multiplier":1,"fmt":"responses","http":"POST","img_in":null,"img_out":null,"in":0,"input":["text"],"label":"OpenAI Responses","max_out":235929,"max_tool_calls":null,"out":0,"output":["text"],"path":"/v1/responses","request_model_id":"xingchen-agi/xing4.0-29b-a4b:free","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]},{"cache_read":null,"cache_write":null,"cache_write_1h":null,"cache_write_5m":null,"conditional_meter":null,"ctx":262144,"family":"openai","fast_multiplier":1,"flex_multiplier":1,"fmt":"chat","http":"POST","img_in":null,"img_out":null,"in":0,"input":["text"],"label":"OpenAI Chat","max_out":235929,"max_tool_calls":null,"out":0,"output":["text"],"path":"/v1/chat/completions","request_model_id":"xingchen-agi/xing4.0-29b-a4b:free","route_label":"Ultra","status_reads":[],"tiers":[],"tool_prices":null,"unit":"token","units":[]}],"model_aliases":[{"canonical_model_id":"xingchen-agi/xing4.0-29b-a4b:free","request_model_id":"xingchen-agi/xing4.0-29b-a4b:free","route_label":"Ultra"}],"name":"XingChen: Xing4.0 29B A4B (free)","price":{"in":0,"in_label":null,"out":0,"out_label":null,"sort_in":0,"sort_out":0,"varies":false},"public_id":"xingchen-agi/xing4.0-29b-a4b:free","short":"China Telecom AI 的 290 億參數文字 MoE 模型，用於規劃、程式碼分析及逐步解決工程問題。"}