xAI Grok 是一款非常强大的 AI 对话系统，只要输入提示词，就能在短短几秒内生成流畅自然的回复。Grok 以其独特的幽默风格和实时网络信息获取能力在业界独树一帜，如今，Grok 已在多个创新领域崭露头角，其影响力正快速扩大。无论是日常对话、创意写作，还是技术分析、代码调试，Grok 都能提供富有洞察力的智能协助，为用户的决策和创作带来全新维度的支持。

本文档主要介绍 Grok Chat Completion API 操作的使用流程，利用它我们可以轻松使用官方 Grok 的对话功能。

申请流程

要使用 Grok Chat Completion API，首先可以到 Grok Chat Completion API 页面点击「Acquire」按钮，获取请求所需要的凭证：

如果你尚未登录或注册，会自动跳转到登录页面邀请您来注册和登录，登录注册之后会自动返回当前页面。

在首次申请时会有免费额度赠送，可以免费使用该 API。

基本使用

接下来就可以在界面上填写对应的内容，如图所示：

在第一次使用该接口时，我们至少需要填写三个内容，一个是 authorization，直接在下拉列表里面选择即可。另一个参数是 model， model 就是我们选择使用 Grok 官网模型类别，这里我们主要有 8 种模型，详情可以看我们提供的模型。最后一个参数是messages，messages是我们输入的提问词数组，它是一个数组，表示可以同时上传多个提问词，每个提问词包含了 role 和 content，其中 role 表示提问者的角色，我们提供了三种身份，分别为 user 、assistant、system 。另一个 content 就是我们提问的具体内容。

同时您可以注意到右侧有对应的调用代码生成，您可以复制代码直接运行，也可以直接点击「Try」按钮进行测试。

调用之后，我们发现返回结果如下：

json { "id": "foaicmpl-13936918-cb99-49e1-b94c-bde98b482ed4", "model": "grok-3", "object": "chat.completion", "created": 1755839683, "choices": [ { "index": 0, "message": { "role": "assistant", "content": "Yo! What's up? 😎 Ready to dive into whatever you're pondering about today?" }, "finish_reason": "stop" } ], "usage": { "prompt_tokens": 8, "completion_tokens": 20, "total_tokens": 28, "prompt_tokens_details": { "cached_tokens": 0, "text_tokens": 0, "audio_tokens": 0, "image_tokens": 0 }, "completion_tokens_details": { "text_tokens": 0, "audio_tokens": 0, "reasoning_tokens": 0 }, "input_tokens": 0, "output_tokens": 0, "input_tokens_details": null } }

返回结果一共有多个字段，介绍如下：

• id，生成此次对话任务的 ID，用于唯一标识此次对话任务。
• model，选择的 Grok 官网模型。
• choices，Grok 针对提问词给于的回答信息。
• usage：针对本次问答对 token 的统计信息。

其中 choices 是包含了 Grok 的回答信息，它里面的 choices 是 Grok回答的具体信息，可以发现如图所示。

可以看到，choices 里面的 content 字段包含了 Grok 回复的具体内容。

流式响应

该接口也支持流式响应，这对网页对接十分有用，可以让网页实现逐字显示效果。

如果想流式返回响应，可以更改请求头里面的 stream 参数，修改为 true。

修改如图所示，不过调用代码需要有对应的更改才能支持流式响应。

将 stream 修改为 true 之后，API 将逐行返回对应的 JSON 数据，在代码层面我们需要做相应的修改来获得逐行的结果。

Python 样例调用代码：

```python import requests

url = "https://api.acedata.cloud/grok/chat/completions"

headers = { "accept": "application/json", "authorization": "Bearer {token}", "content-type": "application/json" }

payload = { "model": "grok-3", "messages": [{"role":"user","content":"Hello"}], "stream": True }

response = requests.post(url, json=payload, headers=headers) print(response.text) ```

输出效果如下：

```json data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"role": "assistant"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "Yo, "}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "what"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "'s g"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "ood?"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": " Rea"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "dy t"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "o di"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "ve i"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "nto "}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "what"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "ever"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": " you"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "'re "}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "pond"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "erin"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "g ab"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "out "}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "toda"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {"content": "y?"}, "logprobs": null, "finish_reason": null, "index": 0}], "usage": null}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": null, "choices": [{"delta": {}, "logprobs": null, "finish_reason": "stop", "index": 0}], "usage": {"prompt_tokens": 0, "completion_tokens": 0, "total_tokens": 0, "prompt_tokens_details": {"cached_tokens": 0, "text_tokens": 0, "audio_tokens": 0, "image_tokens": 0}, "completion_tokens_details": {"text_tokens": 0, "audio_tokens": 0, "reasoning_tokens": 0}, "input_tokens": 0, "output_tokens": 0, "input_tokens_details": null}}

data: {"id": "foaicmpl-503ab14f-3f22-46ab-9f91-3fb44773be38", "object": "chat.completion.chunk", "created": 1755839790, "model": "grok-3", "system_fingerprint": "", "choices": [], "usage": {"prompt_tokens": 8, "completion_tokens": 18, "total_tokens": 26, "prompt_tokens_details": {"cached_tokens": 0, "text_tokens": 0, "audio_tokens": 0, "image_tokens": 0}, "completion_tokens_details": {"text_tokens": 0, "audio_tokens": 0, "reasoning_tokens": 0}, "input_tokens": 0, "output_tokens": 0, "input_tokens_details": null}}

data: [DONE] ```

可以看到，响应里面有许多 data ，data 里面的 choices 即为最新的回答内容，与上文介绍的内容一致。choices 是新增的回答内容，您可以根据结果来对接到您的系统中。同时流式响应的结束是根据 data 的内容来判断的，如果内容为 [DONE]，则表示流式响应回答已经全部结束。返回的 data 结果一共有多个字段，介绍如下：

• id，生成此次对话任务的 ID，用于唯一标识此次对话任务。
• model，选择的 Grok 官网模型。
• choices，Grok 针对提问词给于的回答信息。

JavaScript 也是支持的，比如 Node.js 的流式调用代码如下：

```javascript const options = { method: "post", headers: { "accept": "application/json", "authorization": "Bearer {token}", "content-type": "application/json" }, body: JSON.stringify({ "model": "grok-3", "messages": [{"role":"user","content":"Hello"}], "stream": true }) };

fetch("https://api.acedata.cloud/grok/chat/completions", options) .then(response => response.json()) .then(response => console.log(response)) .catch(err => console.error(err)); ```

Java 样例代码：

```java JSONObject jsonObject = new JSONObject(); jsonObject.put("model", "grok-3"); jsonObject.put("messages", [{"role":"user","content":"Hello"}]); jsonObject.put("stream", true); MediaType mediaType = "application/json; charset=utf-8".toMediaType(); RequestBody body = jsonObject.toString().toRequestBody(mediaType); Request request = new Request.Builder() .url("https://api.acedata.cloud/grok/chat/completions") .post(body) .addHeader("accept", "application/json") .addHeader("authorization", "Bearer {token}") .addHeader("content-type", "application/json") .build();

OkHttpClient client = new OkHttpClient(); Response response = client.newCall(request).execute(); System.out.print(response.body!!.string()) ```

其他语言可以另外自行改写，原理都是一样的。

多轮对话

如果您想要对接多轮对话功能，需要对 messages 字段上传多个提问词，多个提问词的具体示例如下图所示：

Python 样例调用代码：

```python import requests

url = "https://api.acedata.cloud/grok/chat/completions"

headers = { "accept": "application/json", "authorization": "Bearer {token}", "content-type": "application/json" }

payload = { "model": "grok-3", "messages": [{"role":"user","content":"Hello"},{"role":"assistant","content":"What model are you?"},{"role":"user","content":"What did I just say?"}] }

response = requests.post(url, json=payload, headers=headers) print(response.text) ```

通过上传多个提问词，就可以轻松实现多轮对话，可以得到如下回答：

json { "id": "foaicmpl-984ebc53-76b3-4d33-b0e8-0307ab4965af", "model": "grok-3", "object": "chat.completion", "created": 1755839996, "choices": [ { "index": 0, "message": { "role": "assistant", "content": "You said, \"Hello.\"" }, "finish_reason": "stop" } ], "usage": { "prompt_tokens": 27, "completion_tokens": 6, "total_tokens": 33, "prompt_tokens_details": { "cached_tokens": 0, "text_tokens": 0, "audio_tokens": 0, "image_tokens": 0 }, "completion_tokens_details": { "text_tokens": 0, "audio_tokens": 0, "reasoning_tokens": 0 }, "input_tokens": 0, "output_tokens": 0, "input_tokens_details": null } }

可以看到，choices 包含的信息与基本使用的内容是一致的，这个包含了 Grok 针对多个对话进行回复的具体内容，这样就可以根据多个对话内容来回答对应的问题了。

联网实时搜索模型

grok-3-deepsearch 模型与其它模型不同，它可以根据提问词来进行联网实时搜索，并且将联网实时搜索的结果进行适当的调整返回给你，本文将通过一个具体示例来演示联网实时搜索功能，接下来就可以在 Grok Chat Completion API 界面上填写对应的内容，如图所示：

同时您可以注意到右侧有对应的调用代码生成，您可以复制代码直接运行，也可以直接点击「Try」按钮进行测试。

调用之后，我们发现返回结果如下：

json { "id": "chatcmpl-89D08BAul1NoyzuGk017dXkDULBm8", "object": "chat.completion", "created": 1755528106, "model": "grok-3-deepsearch", "choices": [ { "index": 0, "message": { "role": "assistant", "content": "\n> 🔍 **Investigating China news**\n> ***- The request is to find today's news in China.***\n\n> ***- The current date is August 18, 2025, so we'll focus on that.***\n\n> 🔍 **Searching for:** \"latest news China August 18 2025\"\n> 📊 **Results requested:** 15\n\n> 🔧 **Using tool:** x_keyword_search\n> 📝 **Parameters:** query: China filter:news since:2025-08-18 until:2025-08-1..., limit: 20, mode: Latest\n\n> 🔧 **Using tool:** x_semantic_search\n> 📝 **Parameters:** query: breaking news in China today, limit: 15, from_date: 2025-08-18, to_date: 2025-08-19\n\n> ***- Let's search the web for \"latest news China August 18 2025\" to get recent updates.***\n\n> 🔍 **Searching for:** \"latest news China August 18 2025\"\n> 📊 **Results requested:** 15\n\n> 🔧 **Using tool:** x_keyword_search\n> 📝 **Parameters:** query: China filter:news since:2025-08-18 until:2025-08-1..., limit: 20, mode: Latest\n\n> 🔧 **Using tool:** x_semantic_search\n> 📝 **Parameters:** query: breaking news in China today, limit: 15, from_date: 2025-08-18, to_date: 2025-08-19\n\n> *************************************************************************************\n\n\n### Top News Stories from China on August 18, 2025\n\nHere's a summary of the major headlines and developments related to China today, based on recent reports. I've focused on key categories for clarity, drawing from a range of international and domestic sources.\n\n#### International Relations and Diplomacy\n- ******** Reunification Efforts**: China's foreign ministry stated that it is prepared to make a \"great effort\" toward peaceful reunification with*******, emphasizing that the issue is an internal matter. This comes in response to U.S. President Donald Trump's recent comments about discussions with Chinese President Xi Jinping, where Xi reportedly assured no invasion during Trump's term.[[0]](https://www.reuters.com/world/china/china-says-it-is-prepared-great-effort-towards-peaceful-reunification-with-2025-08-18/) [[13]](https://www.usnews.com/news/world/articles/2025-08-18/china-says-it-is-prepared-for-great-effort-towards-peaceful-reunification-with-taiwan) Similar sentiments were echoed in reactions to Trump's diplomacy.[[2]](https://www.newsweek.com/china-trump-putin-zelensky-russia-ukraine-war-2114851)\n- **China-India Talks**: Chinese Foreign Minister Wang Yi arrived in New Delhi for high-level meetings, including with Indian Prime Minister Narendra Modi and External Affairs Minister S. Jaishankar. Discussions stressed border peace, mutual respect, and positive communication amid ongoing tensions. This is Wang's first visit to India in three years, signaling efforts to thaw relations.[[9]](https://www.semafor.com/article/08/18/2025/chinas-top-diplomat-to-meet-indias-modi-in-new-delhi) [[10]](https://www.nytimes.com/2025/08/18/world/asia/india-china-trump.html) Reports suggest U.S. policies under Trump may be inadvertently pushing India closer to China.[[10]](https://www.nytimes.com/2025/08/18/world/asia/india-china-trump.html)\n- **Reactions to Trump-Putin Summit**: China expressed a positive view of the U.S.-Russia talks on Ukraine but warned of increased geopolitical pressure if the U.S. extends NATO-style security guarantees to Ukraine. Analysts in China noted this could complicate Beijing's strategic partnerships with Russia.[[1]](https://www.dw.com/en/trump-putin-meeting-how-is-china-viewing-it/a-73680211) [[2]](https://www.newsweek.com/china-trump-putin-zelensky-russia-ukraine-war-2114851) [[11]](https://www.newsweek.com/thebulletin/2025-08-18)\n- **Other Global Ties**: Germany accused China of threatening stability in the Indo-Pacific and questioning international law, amid broader Western concerns.[[0]](https://www.reuters.com/world/china/china-says-it-is-prepared-great-effort-towards-peaceful-reunification-with-2025-08-18/) (Note: This is from international media; no direct Chinese response reported today.) China also launched a new 18-day Arctic express containership route to Europe, stopping in the UK, Germany, and Poland, highlighting expanding trade infrastructure.[[12]](https://www.highnorthnews.com/en/china-launches-18-day-arctic-express-containership-route-europe-stops-uk-germany-poland)\n\n#### Economy and Business\n- **Stimulus and Consumption Push**: Premier Li Qiang announced China will continue stimulating consumption and protecting livelihoods to sustain economic recovery. This follows positive trends like steady railway freight growth in the first seven months of 2025.[[5]](https://www.reuters.com/markets/asia/china-continue-stimulating-consumption-premier-says-2025-08-18/) [[6]](https://english.www.gov.cn/archive/statistics/202508/18/content_WS68a2f403c6d0868f4e8f4e9a.html) However, analysts noted concerns over China's growth plan \"going cold\" due to incomplete reforms.[[37]](https://x.com/Breakingviews/status/1957340459526771164)\n- **Stock Market Rally**: Chinese stocks gained momentum, with discussions on Bloomberg's \"The China Show\" highlighting market optimism amid global events.[[3]](https://www.bloomberg.com/news/videos/2025-08-18/the-china-show-8-18-2025-video) [[8]](https://www.youtube.com/watch?v=Px3R8Ntev8E)\n- **Trade and Tariffs**: The U.S. and China extended a tariff truce by 90 days, avoiding a surge in duties. Negotiations for a broader trade deal remain challenging.[[7]](https://issi.org.pk/cpsc-monitor-august-18-2025/) EV producers are leading China's \"going global\" strategy, exploring offshore economic expansion.[[23]](https://x.com/shehzadyounis/status/1957451550106001585)\n- **AI and Tech**: Trump's proposed AI plan is positioned as a counter to China's rising tech influence, per U.S. opinion pieces.[[30]](https://x.com/FoxNewsAI/status/1957449958535139450)\n\n#### Domestic Developments\n- **Infrastructure and Growth**: China reported efficient railway operations and record-breaking projects, like the Huajiang Canyon Bridge, completed in three years as the world's tallest.[[6]](https://english.www.gov.cn/archive/statistics/202508/18/content_WS68a2f403c6d0868f4e8f4e9a.html) [[15]](https://x.com/AxorReturn/status/1957435608898396505)\n- **\n---\nMessage contains too many sensitive words!-Related Warnings**: Taipei" }, "finish_reason": "stop" } ], "usage": { "prompt_tokens": 15, "completion_tokens": 1618, "total_tokens": 1633, "prompt_tokens_details": { "text_tokens": 8 }, "completion_tokens_details": { "content_tokens": 1618 } } }

可以看到，choices 里面的回答信息是根据联网实时搜索后得到的，并且也给出了相关的链接，choices 里面的回答信息是要通过 markdown 语法进行渲染，这样才能获得最佳的体验，最后这也体现出我们模型的联网功能的强大优势。

错误处理

在调用 API 时，如果遇到错误，API 会返回相应的错误代码和信息。例如：

• 400 token_mismatched：Bad request, possibly due to missing or invalid parameters.
• 400 api_not_implemented：Bad request, possibly due to missing or invalid parameters.
• 401 invalid_token：Unauthorized, invalid or missing authorization token.
• 429 too_many_requests：Too many requests, you have exceeded the rate limit.
• 500 api_error：Internal server error, something went wrong on the server.

错误响应示例

{ "success": false, "error": { "code": "api_error", "message": "fetch failed" }, "trace_id": "2cf86e86-22a4-46e1-ac2f-032c0f2a4e89" }

结论

通过本文档，您已经了解了如何使用 OpenAI Chat Completion API 轻松实现官方 OpenAI ChatGPT 的对话功能。希望本文档能帮助您更好地对接和使用该 API。如有任何问题，请随时联系我们的技术支持团队。