> 如果你是 Apple IAP 账号，请联系我将你的账号转为邮箱账号
> If you are an Apple IAP account, please contact me to convert your account to an email account
Note: We have migrated AI's Chat UI to Zhi
提示：我们已经将 AI 的 Chat UI 迁移到 Zhi
- Download and Install Zhi.
- Signin with your email.
- In the upper right corner, click the + icon
- Select
Join Chat - Type
{{chat.ChatUUID}}:{{chat.Key}}in Chat and Key - Type any string in Chat Name and choose Chat Avatar
- Type any string in Your Name and choose Your Avatar
- Tap Join button
- 下载安装 Zhi.
- 用邮箱登录
- 在右上角点击 + 图标
- 选择
加入 Chat - 输入
{{chat.ChatUUID}}:{{chat.Key}}在 Chat and Key - 输入 Chat Name 和选择一个 Chat 头像
- 输入你的昵称和选择一个你的头像
- 点击加入按钮
{{ zh ? '价格' : 'Price' }}
| Model | Input | Output | Context |
|---|---|---|---|
| GPT-3.5-TURBO | $0.0015 / 1k tokens | $0.003 / 1k tokens | Yes |
| GPT-4-TURBO | $0.015 / 1k tokens | $0.045 / 1k tokens | Yes |
| GPT-4 | $0.045 / 1k tokens | $0.09 / 1k tokens | Yes |
Why can't I receive the verification code by email?
Many people often have their email address typed wrong, remember to check carefully. Also don't forget to check the spam mailbox. If you still cannot receive the verification code after 5 minutes, it may be that your network cannot access our API server. You can try another network. If you are using Microsoft's mailbox, users around the world generally report that they will lose emails without any notification, stay away.
为什么邮箱收不到验证码
很多人会输错邮件地址，记得仔细检查下，同时不要忘记检查垃圾箱。如果五分钟仍然收不到验证码，可能是你的网络无法访问我们的 API 服务器，可以尝试换个网络，比如用 Wi-Fi 或 4G。如果你正在用微软的邮箱，全球用户普遍反馈会无任何通知丢邮件，远离。
Token length of models
Each model has a maximum token length, and we know the model's memory capacity because we provide the previous conversation as a prompt to the model. Therefore, token length directly affects the context memory capacity and maximum character length. Currently, GPT-3.5-TURBO is 16k, GPT-4-TURBO is 128k, while all other models have a limit of 8k tokens.
模型的 Token 长度
每个模型都有最大 Token 长度，我们知道模型的记忆能力是因为我们将之前的会话内容又作为 Prompt 发给了模型，所以 Token 长度直接影响上下文记忆能力及最大处理字符长度。目前 GPT-3.5-TURBO 为 16k，GPT-4-TURBO 为 128k, 其余模型均为 8k。
Context
GPT-3.5-TURBO, GPT-4, GPT-4-TURBO support context, which means they will remember the content of the current conversation.
上下文
GPT-3.5-TURBO, GPT-4, GPT-4-TURBO 支持上下文，意思就是会在记忆当前会话上文的内容。