type
status
date
slug
summary
tags
category
icon
类型
标签
参考标签
在这篇帖子里,介绍了AIHUBMIX的API,因为他们充值有折扣,充1美元只需6.3元,相对1美元7.3的实际费率,折扣为86折,所以他们的API还是很划算的,用了这么久,响应速度也很快,服务稳定。
现在用API最主要的地方就是Glarity,总结检索结果,是一个很高频的应用场景。之前使用的模型是GPT-3.5。最近AIHUBMIX终于公布了所有模型的费率,发现文心一言的价格非常便宜,所以就换了。效果也很不错。
价格按照输出费率排序如下:
模型 | 输入$/1M tokens | 输出$/1M tokens |
yi-34b-chat-0205 | 0.000355 | 0.000355 |
yi-vl-plus | 0.000852 | 0.000852 |
yi-34b-chat-200k | 0.001704 | 0.001704 |
text-embedding-3-small | 0.020000 | 0.020000 |
embedding-2 | 0.071000 | 0.071000 |
Qwen/Qwen2-7B-Instruct | 0.080000 | 0.080000 |
ahm-Phi-3-medium-4k | 0.100000 | 0.100000 |
ahm-Phi-3-small-128k | 0.100000 | 0.100000 |
gemma-7b-it | 0.100000 | 0.100000 |
gemma2-9b-it | 0.100000 | 0.100000 |
text-embedding-ada-002 | 0.100000 | 0.100000 |
gemini-1.5-flash | 0.400000 | 0.112000 |
llama3-8b-8192 | 0.060000 | 0.120000 |
text-embedding-3-large | 0.130000 | 0.130000 |
text-moderation-latest | 0.200000 | 0.200000 |
text-moderation-stable | 0.200000 | 0.200000 |
Qwen/Qwen2-57B-A14B-Instruct | 0.240000 | 0.240000 |
deepseek-ai/deepseek-v2-chat | 0.142000 | 0.284000 |
qwen-long | 0.080000 | 0.320000 |
deepseek-chat | 0.160000 | 0.320000 |
deepseek-coder | 0.160000 | 0.320000 |
babbage-002 | 0.400000 | 0.400000 |
text-ada-001 | 0.400000 | 0.400000 |
yi-medium | 0.400000 | 0.400000 |
mixtral-8x7b-32768 | 0.500000 | 0.500000 |
text-babbage-001 | 0.500000 | 0.500000 |
gemini-pro | 0.200000 | 0.600000 |
aihubmix-Llama-3-70B-Instruct | 0.700000 | 0.700000 |
glm-3-turbo | 0.710000 | 0.710000 |
Qwen/Qwen2-72B-Instruct | 0.800000 | 0.800000 |
llama3-70b-8192 | 0.700000 | 0.937288 |
qwen-turbo | 0.320000 | 0.960000 |
claude-3-haiku-20240307 | 0.260000 | 1.300000 |
gpt-3.5-turbo-0125 | 0.500000 | 1.500000 |
gpt-3.5-turbo | 0.500000 | 1.500000 |
moonshot-v1-8k | 1.704000 | 1.704000 |
qwen-plus | 0.600000 | 1.800000 |
yi-large-turbo | 1.800000 | 1.800000 |
command-r | 0.640000 | 1.920000 |
gpt-3.5-turbo-1106 | 1.000000 | 2.000000 |
command | 1.000000 | 2.000000 |
command-light | 1.000000 | 2.000000 |
command-light-nightly | 1.000000 | 2.000000 |
command-nightly | 1.000000 | 2.000000 |
gpt-3.5-turbo-instruct | 1.500000 | 2.000000 |
gpt-3.5-turbo-0613 | 1.500000 | 2.000000 |
gpt-3.5-turbo-0301 | 1.500000 | 2.000000 |
davinci-002 | 2.000000 | 2.000000 |
text-curie-001 | 2.000000 | 2.000000 |
gemini-pro-vision | 1.000000 | 3.000000 |
yi-large | 3.000000 | 3.000000 |
deepseek-ai/deepseek-llm-67b-chat | 1.600000 | 3.200000 |
moonshot-v1-32k | 3.408000 | 3.408000 |
claude-instant-1.2 | 0.800000 | 3.600000 |
gpt-3.5-turbo-16k-0613 | 3.000000 | 4.000000 |
gpt-3.5-turbo-16k | 3.000000 | 4.000000 |
yi-large-rag | 4.000000 | 4.000000 |
moonshot-v1-128k | 8.520000 | 8.520000 |
Mistral-large-hwlia | 4.000000 | 12.000000 |
aihubmix-Mistral-large | 4.000000 | 12.000000 |
mistral-large-latest | 4.000000 | 12.000000 |
glm-4 | 14.200000 | 14.200000 |
glm-4v | 14.200000 | 14.200000 |
claude-3-sonnet-20240229 | 3.000000 | 15.000000 |
claude-3-5-sonnet-20240620 | 3.000000 | 15.000000 |
gpt-4o-2024-05-13 | 5.000000 | 15.000000 |
gpt-4o | 5.000000 | 15.000000 |
tts-1 | 15.000000 | 15.000000 |
dall-e-2 | 16.000000 | 16.000000 |
qwen-max | 6.000000 | 18.000000 |
qwen-max-longcontext | 6.000000 | 18.000000 |
aihubmix-command-r-plus | 3.840000 | 19.200000 |
command-r-plus | 3.840000 | 19.200000 |
text-davinci-002 | 20.000000 | 20.000000 |
text-davinci-003 | 20.000000 | 20.000000 |
text-davinci-edit-001 | 20.000000 | 20.000000 |
gemini-1.5-pro | 8.000000 | 24.000000 |
gpt-4-vision-preview | 10.000000 | 30.000000 |
gpt-4-turbo-preview | 10.000000 | 30.000000 |
gpt-4-turbo-2024-04-09 | 10.000000 | 30.000000 |
gpt-4-turbo | 10.000000 | 30.000000 |
gpt-4-1106-preview | 10.000000 | 30.000000 |
gpt-4-0125-preview | 10.000000 | 30.000000 |
tts-1-hd | 30.000000 | 30.000000 |
dall-e-3 | 40.000000 | 40.000000 |
claude-2.1 | 11.020000 | 49.590000 |
claude-2.0 | 11.020000 | 49.590000 |
gpt-4-0613 | 30.000000 | 60.000000 |
gpt-4-0314 | 30.000000 | 60.000000 |
gpt-4 | 30.000000 | 60.000000 |
Qwen/Qwen2-1.5B-Instruct | 60.000000 | 60.000000 |
THUDM/glm-4-9b-chat | 60.000000 | 60.000000 |
accounts/fireworks/models/deepseek-coder-v2-lite-instruct | 60.000000 | 60.000000 |
meta/llama3-8B-chat | 60.000000 | 60.000000 |
mixtralai/Mixtral-8x22B-Instruct-v0.1 | 60.000000 | 60.000000 |
whisper-1 | 60.000000 | 60.000000 |
whisper-large-v3 | 60.000000 | 60.000000 |
claude-3-opus-20240229 | 15.000000 | 75.000000 |
gpt-4-32k-0613 | 60.000000 | 120.000000 |
gpt-4-32k-0314 | 60.000000 | 120.000000 |
gpt-4-32k | 60.000000 | 120.000000 |
deepseek-ai/DeepSeek-Coder-V2-Instruct | 60.000000 | 120.000000 |
gpt-4-1106-vision-preview | 60.000000 | 180.000000 |
- 作者:Neo Zed
- 链接:https://musingpages.com/AI/2024/07/11/api-rates
- 声明:本文采用 CC BY-NC-SA 4.0 许可协议,转载请注明出处。