OpenAI

价格

Service	Input	Output	Notes
gpt-4-turbo 128k	$10.00 /1M tokens	$30.00 /1M tokens	更好更便宜
gpt-4	$30.00 /1M tokens	$60.00 /1M tokens
gpt-4-32k	$60.00 /1M tokens	$120.00 /1M tokens
gpt-3.5-turbo-0125	$0.50 /1M tokens	$1.50 /1M tokens
gpt-3.5-turbo-instruct	$1.50 /1M tokens	$2.00 /1M tokens

• 1M tokens 大约 50 万字

---

• InstructGPT - 单个指令生成
  • Ada - $0.0004 / 1K tokens
    • 最快
    • 可以 Embedding
  • Babbage - $0.0005 / 1K tokens
  • Curie - $0.002 / 1K tokens
  • Davinci - $0.02 / 1K tokens
    • 最强
• Fine-tuning InstructGPT
  • 训练价格 ~= 基础模型
  • 使用价格 = 训练价格 x 4
• DALL·E
  • 1024×1024 $0.02/image
  • 512×512 $0.018/image
  • 256×256 $0.016/image
• Audio
  • Whisper $0.006 / minute
• https://openai.com/pricing

Models

• gpt-3.5-turbo-0613
  • 16K 上下文

API

• https://platform.openai.com/docs/api-reference
• https://platform.openai.com/docs/models/model-endpoint-compatibility
• https://npmjs.com/package/chatgpt

curl https://api.openai.com/v1/models \
  -H "Authorization: Bearer $OPENAI_API_KEY" \
  -H "OpenAI-Organization: YOUR_ORG_ID"

{
  "error": {
    "message": "This key is associated with a deactivated account. If you feel this is an error, contact us through our help center at help.openai.com.",
    "type": "invalid_request_error",
    "param": null,
    "code": "account_deactivated"
  }
}

限制

-	Text&Embedding	Chat	Codex	Edit	Image	Audio
Pay-as-you-go 48h	60 RPM, 250,000 TPM	60 RPM, 60,000 TPM	20 RPM , 40,000 TPM	20 RPM , 150,000 TPM	50 images / min	50 RPM
Pay-as-you-go	3,500 RPM,350,000 TPM	3,500 RPM, 90,000 TPM	20 RPM, 40,000 TPM	20 RPM ,150,000 TPM	50 images / min	50 RPM

• TPM - tokens per minute - Token/分钟
• RPM - requests per minute - 请求/分钟
• 限制是组织维度
• 替代方案
  • Azure OpenAI
  • 多组织
• 1 token ~= 0.5-1 汉字, 1 token ~= 4 字母
• avg 15 token/s, min 0.7 token/s, max 50 token/s - 基于历史消息统计
• 3,500 RPM, 90,000 TPM - 1 个账号，两种维度限制，请求 和 Token
  • 3500/60 = 58.3 RPS ~ 60/s 并发请求
  • 90000/60 = 1500 TPS
    • 1500/50 = 30/s 并发请求 - 回复内容多
    • 1500/15 = 100/s 并发请求
    • 1500/0.7 = 2142/s 并发请求 - 回复内容少
• 限流都是分钟级别
  • 也就是说持续 1 分钟 RPS 达到 60 就会有问题 - 例如: 1 分钟有 3600 人操作，每秒操作 1。次
  • 基于 Token 的限制则影响有多少个持续的回答

TYPE	1 TPM EQUALS
davinci	1 token per minute
curie	25 tokens per minute
babbage	100 tokens per minute
ada	200 tokens per minute

• https://platform.openai.com/docs/guides/rate-limits

fin tune

pip install --upgrade openai
export OPENAI_API_KEY="<OPENAI_API_KEY>"
# openai tools fine_tunes.prepare_data -f <LOCAL_FILE>

# ada, babbage, curie, davinci
openai api fine_tunes.create -t <TRAIN_FILE_ID_OR_PATH> -m <BASE_MODEL>
openai api fine_tunes.follow -i <YOUR_FINE_TUNE_JOB_ID>

Rate limit reached for default-gpt-3.5-turbo in organization org-xxx on requests per min. Limit: 60 / min.

Rate limit reached for default-gpt-3.5-turbo in organization org-xxx on requests per min. Limit: 60 / min. Please try again in 1s. Contact us through our help center at help.openai.com if you continue to have issues.

您的银行卡被拒绝了。

Access denied

尝试不用 VPN 或者换 VPN

清理 Auth 信息

javascript:window.localStorage.removeItem(Object.keys(window.localStorage).find(i=>i.startsWith('@@auth0spajs'))