xinyin025/one-api
1
0
Fork 0
mirror of https://github.com/songquanpeng/one-api.git synced 2026-02-04 21:25:56 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge remote-tracking branch 'remotes/origin_songquanpeng/main'

Browse Source 
# Conflicts:
#	relay/adaptor/openai/adaptor.go
#	relay/controller/text.go
This commit is contained in:
mlkt
2024-08-07 06:24:20 +08:00
parent c885953c6d f9774698e9
commit 04aee84485
114 changed files with 3392 additions and 1700 deletions
Download Patch File Download Diff File Expand all files Collapse all files

108
relay/billing/ratio/model.go
View File

@@ -2,6 +2,7 @@ package ratio
import (
"encoding/json"
"fmt"
"strings"
"github.com/songquanpeng/one-api/common/logger"
@@ -27,15 +28,17 @@ var ModelRatio = map[string]float64{
"gpt-4-32k": 30,
"gpt-4-32k-0314": 30,
"gpt-4-32k-0613": 30,
"gpt-4-1106-preview": 5, // $0.01 / 1K tokens
"gpt-4-0125-preview": 5, // $0.01 / 1K tokens
"gpt-4-turbo-preview": 5, // $0.01 / 1K tokens
"gpt-4-turbo": 5, // $0.01 / 1K tokens
"gpt-4-turbo-2024-04-09": 5, // $0.01 / 1K tokens
"gpt-4o": 2.5, // $0.005 / 1K tokens
"gpt-4o-2024-05-13": 2.5, // $0.005 / 1K tokens
"gpt-4-vision-preview": 5, // $0.01 / 1K tokens
"gpt-3.5-turbo": 0.25, // $0.0005 / 1K tokens
"gpt-4-1106-preview": 5, // $0.01 / 1K tokens
"gpt-4-0125-preview": 5, // $0.01 / 1K tokens
"gpt-4-turbo-preview": 5, // $0.01 / 1K tokens
"gpt-4-turbo": 5, // $0.01 / 1K tokens
"gpt-4-turbo-2024-04-09": 5, // $0.01 / 1K tokens
"gpt-4o": 2.5, // $0.005 / 1K tokens
"gpt-4o-2024-05-13": 2.5, // $0.005 / 1K tokens
"gpt-4o-mini": 0.075, // $0.00015 / 1K tokens
"gpt-4o-mini-2024-07-18": 0.075, // $0.00015 / 1K tokens
"gpt-4-vision-preview": 5, // $0.01 / 1K tokens
"gpt-3.5-turbo": 0.25, // $0.0005 / 1K tokens
"gpt-3.5-turbo-0301": 0.75,
"gpt-3.5-turbo-0613": 0.75,
"gpt-3.5-turbo-16k": 1.5, // $0.003 / 1K tokens
@@ -70,12 +73,13 @@ var ModelRatio = map[string]float64{
"dall-e-2": 0.02 * USD, // $0.016 - $0.020 / image
"dall-e-3": 0.04 * USD, // $0.040 - $0.120 / image
// https://www.anthropic.com/api#pricing
"claude-instant-1.2": 0.8 / 1000 * USD,
"claude-2.0": 8.0 / 1000 * USD,
"claude-2.1": 8.0 / 1000 * USD,
"claude-3-haiku-20240307": 0.25 / 1000 * USD,
"claude-3-sonnet-20240229": 3.0 / 1000 * USD,
"claude-3-opus-20240229": 15.0 / 1000 * USD,
"claude-instant-1.2": 0.8 / 1000 * USD,
"claude-2.0": 8.0 / 1000 * USD,
"claude-2.1": 8.0 / 1000 * USD,
"claude-3-haiku-20240307": 0.25 / 1000 * USD,
"claude-3-sonnet-20240229": 3.0 / 1000 * USD,
"claude-3-5-sonnet-20240620": 3.0 / 1000 * USD,
"claude-3-opus-20240229": 15.0 / 1000 * USD,
// https://cloud.baidu.com/doc/WENXINWORKSHOP/s/hlrk4akp7
"ERNIE-4.0-8K": 0.120 * RMB,
"ERNIE-3.5-8K": 0.012 * RMB,
@@ -94,12 +98,11 @@ var ModelRatio = map[string]float64{
"bge-large-en": 0.002 * RMB,
"tao-8k": 0.002 * RMB,
// https://ai.google.dev/pricing
"PaLM-2": 1,
"gemini-pro": 1, // $0.00025 / 1k characters -> $0.001 / 1k tokens
"gemini-pro-vision": 1, // $0.00025 / 1k characters -> $0.001 / 1k tokens
"gemini-1.0-pro-vision-001": 1,
"gemini-1.0-pro-001": 1,
"gemini-1.5-pro": 1,
"gemini-pro": 1, // $0.00025 / 1k characters -> $0.001 / 1k tokens
"gemini-1.0-pro": 1,
"gemini-1.5-flash": 1,
"gemini-1.5-pro": 1,
"aqa": 1,
// https://open.bigmodel.cn/pricing
"glm-4": 0.1 * RMB,
"glm-4v": 0.1 * RMB,
@@ -124,6 +127,7 @@ var ModelRatio = map[string]float64{
"SparkDesk-v2.1": 1.2858, // ￥0.018 / 1k tokens
"SparkDesk-v3.1": 1.2858, // ￥0.018 / 1k tokens
"SparkDesk-v3.5": 1.2858, // ￥0.018 / 1k tokens
"SparkDesk-v4.0": 1.2858, // ￥0.018 / 1k tokens
"360GPT_S2_V9": 0.8572, // ¥0.012 / 1k tokens
"embedding-bert-512-v1": 0.0715, // ¥0.001 / 1k tokens
"embedding_s1_v1": 0.0715, // ¥0.001 / 1k tokens
@@ -153,12 +157,16 @@ var ModelRatio = map[string]float64{
"mistral-large-latest": 8.0 / 1000 * USD,
"mistral-embed": 0.1 / 1000 * USD,
// https://wow.groq.com/#:~:text=inquiries%C2%A0here.-,Model,-Current%20Speed
"llama3-70b-8192": 0.59 / 1000 * USD,
"mixtral-8x7b-32768": 0.27 / 1000 * USD,
"llama3-8b-8192": 0.05 / 1000 * USD,
"gemma-7b-it": 0.1 / 1000 * USD,
"llama2-70b-4096": 0.64 / 1000 * USD,
"llama2-7b-2048": 0.1 / 1000 * USD,
"gemma-7b-it": 0.07 / 1000000 * USD,
"mixtral-8x7b-32768": 0.24 / 1000000 * USD,
"llama3-8b-8192": 0.05 / 1000000 * USD,
"llama3-70b-8192": 0.59 / 1000000 * USD,
"gemma2-9b-it": 0.20 / 1000000 * USD,
"llama-3.1-405b-reasoning": 0.89 / 1000000 * USD,
"llama-3.1-70b-versatile": 0.59 / 1000000 * USD,
"llama-3.1-8b-instant": 0.05 / 1000000 * USD,
"llama3-groq-70b-8192-tool-use-preview": 0.89 / 1000000 * USD,
"llama3-groq-8b-8192-tool-use-preview": 0.19 / 1000000 * USD,
// https://platform.lingyiwanwu.com/docs#-计费单元
"yi-34b-chat-0205": 2.5 / 1000 * RMB,
"yi-34b-chat-200k": 12.0 / 1000 * RMB,
@@ -167,6 +175,9 @@ var ModelRatio = map[string]float64{
"step-1v-32k": 0.024 * RMB,
"step-1-32k": 0.024 * RMB,
"step-1-200k": 0.15 * RMB,
// aws llama3 https://aws.amazon.com/cn/bedrock/pricing/
"llama3-8b-8192(33)": 0.0003 / 0.002, // $0.0003 / 1K tokens
"llama3-70b-8192(33)": 0.00265 / 0.002, // $0.00265 / 1K tokens
// https://cohere.com/pricing
"command": 0.5,
"command-nightly": 0.5,
@@ -183,7 +194,11 @@ var ModelRatio = map[string]float64{
"deepl-ja": 25.0 / 1000 * USD,
}
var CompletionRatio = map[string]float64{}
var CompletionRatio = map[string]float64{
// aws llama3
"llama3-8b-8192(33)": 0.0006 / 0.0003,
"llama3-70b-8192(33)": 0.0035 / 0.00265,
}
var DefaultModelRatio map[string]float64
var DefaultCompletionRatio map[string]float64
@@ -232,22 +247,28 @@ func UpdateModelRatioByJSONString(jsonStr string) error {
return json.Unmarshal([]byte(jsonStr), &ModelRatio)
}
func GetModelRatio(name string) float64 {
func GetModelRatio(name string, channelType int) float64 {
if strings.HasPrefix(name, "qwen-") && strings.HasSuffix(name, "-internet") {
name = strings.TrimSuffix(name, "-internet")
}
if strings.HasPrefix(name, "command-") && strings.HasSuffix(name, "-internet") {
name = strings.TrimSuffix(name, "-internet")
}
ratio, ok := ModelRatio[name]
if !ok {
ratio, ok = DefaultModelRatio[name]
model := fmt.Sprintf("%s(%d)", name, channelType)
if ratio, ok := ModelRatio[model]; ok {
return ratio
}
if !ok {
logger.SysError("model ratio not found: " + name)
return 30
if ratio, ok := DefaultModelRatio[model]; ok {
return ratio
}
return ratio
if ratio, ok := ModelRatio[name]; ok {
return ratio
}
if ratio, ok := DefaultModelRatio[name]; ok {
return ratio
}
logger.SysError("model ratio not found: " + name)
return 30
}
func CompletionRatio2JSONString() string {
@@ -263,7 +284,17 @@ func UpdateCompletionRatioByJSONString(jsonStr string) error {
return json.Unmarshal([]byte(jsonStr), &CompletionRatio)
}
func GetCompletionRatio(name string) float64 {
func GetCompletionRatio(name string, channelType int) float64 {
if strings.HasPrefix(name, "qwen-") && strings.HasSuffix(name, "-internet") {
name = strings.TrimSuffix(name, "-internet")
}
model := fmt.Sprintf("%s(%d)", name, channelType)
if ratio, ok := CompletionRatio[model]; ok {
return ratio
}
if ratio, ok := DefaultCompletionRatio[model]; ok {
return ratio
}
if ratio, ok := CompletionRatio[name]; ok {
return ratio
}
@@ -282,6 +313,9 @@ func GetCompletionRatio(name string) float64 {
return 4.0 / 3.0
}
if strings.HasPrefix(name, "gpt-4") {
if strings.HasPrefix(name, "gpt-4o-mini") {
return 4
}
if strings.HasPrefix(name, "gpt-4-turbo") ||
strings.HasPrefix(name, "gpt-4o") ||
strings.HasSuffix(name, "preview") {