mirror of
https://github.com/lobehub/lobe-chat.git
synced 2026-06-15 04:00:09 +00:00
💄 style: update Wenxin & Hunyuan model list (#6995)
This commit is contained in:
+131
-109
@@ -49,7 +49,6 @@ const hunyuanChatModels: AIChatModelCard[] = [
|
||||
description:
|
||||
'采用更优的路由策略,同时缓解了负载均衡和专家趋同的问题。长文方面,大海捞针指标达到99.9%。MOE-256K 在长度和效果上进一步突破,极大的扩展了可输入长度。',
|
||||
displayName: 'Hunyuan Standard 256K',
|
||||
enabled: true,
|
||||
id: 'hunyuan-standard-256K',
|
||||
maxOutput: 6000,
|
||||
pricing: {
|
||||
@@ -63,95 +62,6 @@ const hunyuanChatModels: AIChatModelCard[] = [
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 32_000,
|
||||
description:
|
||||
'通用体验优化,包括NLP理解、文本创作、闲聊、知识问答、翻译、领域等;提升拟人性,优化模型情商;提升意图模糊时模型主动澄清能力;提升字词解析类问题的处理能力;提升创作的质量和可互动性;提升多轮体验。',
|
||||
displayName: 'Hunyuan Turbo Latest',
|
||||
enabled: true,
|
||||
id: 'hunyuan-turbo-latest',
|
||||
maxOutput: 4000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 15,
|
||||
output: 50,
|
||||
},
|
||||
releasedAt: '2025-01-10',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 32_000,
|
||||
description:
|
||||
'本版本优化:数据指令scaling,大幅提升模型通用泛化能力;大幅提升数学、代码、逻辑推理能力;优化文本理解字词理解相关能力;优化文本创作内容生成质量',
|
||||
displayName: 'Hunyuan Turbo',
|
||||
id: 'hunyuan-turbo',
|
||||
maxOutput: 4000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 15,
|
||||
output: 50,
|
||||
},
|
||||
releasedAt: '2025-01-10',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 32_000,
|
||||
description:
|
||||
'本版本优化:数据指令scaling,大幅提升模型通用泛化能力;大幅提升数学、代码、逻辑推理能力;优化文本理解字词理解相关能力;优化文本创作内容生成质量',
|
||||
displayName: 'Hunyuan Turbo 20241223',
|
||||
id: 'hunyuan-turbo-20241223',
|
||||
maxOutput: 4000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 15,
|
||||
output: 50,
|
||||
},
|
||||
releasedAt: '2025-01-10',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 32_000,
|
||||
description:
|
||||
'hunyuan-turbo 2024 年 11 月 20 日固定版本,介于 hunyuan-turbo 和 hunyuan-turbo-latest 之间的一个版本。',
|
||||
displayName: 'Hunyuan Turbo 20241120',
|
||||
id: 'hunyuan-turbo-20241120',
|
||||
maxOutput: 4000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 15,
|
||||
output: 50,
|
||||
},
|
||||
releasedAt: '2024-11-20',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
search: true,
|
||||
@@ -196,6 +106,118 @@ const hunyuanChatModels: AIChatModelCard[] = [
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 32_000,
|
||||
description:
|
||||
'通用体验优化,包括NLP理解、文本创作、闲聊、知识问答、翻译、领域等;提升拟人性,优化模型情商;提升意图模糊时模型主动澄清能力;提升字词解析类问题的处理能力;提升创作的质量和可互动性;提升多轮体验。',
|
||||
displayName: 'Hunyuan Turbo',
|
||||
enabled: true,
|
||||
id: 'hunyuan-turbo-latest',
|
||||
maxOutput: 4000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 2.4,
|
||||
output: 9.6,
|
||||
},
|
||||
releasedAt: '2025-01-10',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 32_000,
|
||||
description:
|
||||
'本版本优化:数据指令scaling,大幅提升模型通用泛化能力;大幅提升数学、代码、逻辑推理能力;优化文本理解字词理解相关能力;优化文本创作内容生成质量',
|
||||
displayName: 'Hunyuan Turbo 20241223',
|
||||
id: 'hunyuan-turbo-20241223',
|
||||
maxOutput: 4000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 2.4,
|
||||
output: 9.6,
|
||||
},
|
||||
releasedAt: '2025-01-10',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 32_000,
|
||||
description:
|
||||
'hunyuan-TurboS 混元旗舰大模型最新版本,具备更强的思考能力,更优的体验效果。',
|
||||
displayName: 'Hunyuan TurboS',
|
||||
enabled: true,
|
||||
id: 'hunyuan-turbos-latest',
|
||||
maxOutput: 8000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.8,
|
||||
output: 2,
|
||||
},
|
||||
releasedAt: '2025-03-13',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 32_000,
|
||||
description:
|
||||
'统一数学解题步骤的风格,加强数学多轮问答。文本创作优化回答风格,去除AI味,增加文采。',
|
||||
displayName: 'Hunyuan TurboS 20250313',
|
||||
id: 'hunyuan-turbos-20250313',
|
||||
maxOutput: 8000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.8,
|
||||
output: 2,
|
||||
},
|
||||
releasedAt: '2025-03-13',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 32_000,
|
||||
description:
|
||||
'hunyuan-TurboS pv2.1.2 固定版本预训练底座训练token 数升级;数学/逻辑/代码等思考能力提升;中英文通用体验效果提升,包括文本创作、文本理解、知识问答、闲聊等。',
|
||||
displayName: 'Hunyuan TurboS 20250226',
|
||||
id: 'hunyuan-turbos-20250226',
|
||||
maxOutput: 8000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.8,
|
||||
output: 2,
|
||||
},
|
||||
releasedAt: '2025-02-25',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
vision: true,
|
||||
@@ -267,8 +289,8 @@ const hunyuanChatModels: AIChatModelCard[] = [
|
||||
maxOutput: 4000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 4,
|
||||
output: 8,
|
||||
input: 3.5,
|
||||
output: 7,
|
||||
},
|
||||
releasedAt: '2024-11-12',
|
||||
type: 'chat',
|
||||
@@ -306,21 +328,6 @@ const hunyuanChatModels: AIChatModelCard[] = [
|
||||
releasedAt: '2024-07-04',
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
contextWindowTokens: 8000,
|
||||
description:
|
||||
'支持中文和英语、日语、法语、葡萄牙语、西班牙语、土耳其语、俄语、阿拉伯语、韩语、意大利语、德语、越南语、马来语、印尼语15种语言互译,基于多场景翻译评测集自动化评估COMET评分,在十余种常用语种中外互译能力上整体优于市场同规模模型。',
|
||||
displayName: 'Hunyuan Translation',
|
||||
id: 'hunyuan-translation',
|
||||
maxOutput: 4000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 25,
|
||||
output: 75,
|
||||
},
|
||||
releasedAt: '2024-10-25',
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
contextWindowTokens: 8000,
|
||||
description:
|
||||
@@ -330,12 +337,27 @@ const hunyuanChatModels: AIChatModelCard[] = [
|
||||
maxOutput: 4000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 5,
|
||||
output: 15,
|
||||
input: 1,
|
||||
output: 3,
|
||||
},
|
||||
releasedAt: '2024-11-25',
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
contextWindowTokens: 8000,
|
||||
description:
|
||||
'支持中文和英语、日语、法语、葡萄牙语、西班牙语、土耳其语、俄语、阿拉伯语、韩语、意大利语、德语、越南语、马来语、印尼语15种语言互译,基于多场景翻译评测集自动化评估COMET评分,在十余种常用语种中外互译能力上整体优于市场同规模模型。',
|
||||
displayName: 'Hunyuan Translation',
|
||||
id: 'hunyuan-translation',
|
||||
maxOutput: 4000,
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 15,
|
||||
output: 45,
|
||||
},
|
||||
releasedAt: '2024-10-25',
|
||||
type: 'chat',
|
||||
},
|
||||
];
|
||||
|
||||
export const allModels = [...hunyuanChatModels];
|
||||
|
||||
+221
-33
@@ -1,63 +1,47 @@
|
||||
import { AIChatModelCard } from '@/types/aiModel';
|
||||
|
||||
const wenxinChatModels: AIChatModelCard[] = [
|
||||
/*
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
resoning: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 8192,
|
||||
contextWindowTokens: 32_768,
|
||||
description:
|
||||
'百度自研的旗舰级大规模⼤语⾔模型,覆盖海量中英文语料,具有强大的通用能力,可满足绝大部分对话问答、创作生成、插件应用场景要求;支持自动对接百度搜索插件,保障问答信息时效。',
|
||||
displayName: 'ERNIE 3.5 8K',
|
||||
'文心大模型X1具备更强的理解、规划、反思、进化能力。作为能力更全面的深度思考模型,文心X1兼备准确、创意和文采,在中文知识问答、文学创作、文稿写作、日常对话、逻辑推理、复杂计算及工具调用等方面表现尤为出色。',
|
||||
displayName: 'ERNIE X1 32K Preview',
|
||||
enabled: true,
|
||||
id: 'ernie-3.5-8k',
|
||||
id: 'ernie-x1-32k-preview',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.8,
|
||||
output: 2,
|
||||
input: 2,
|
||||
output: 8,
|
||||
},
|
||||
releasedAt: '2025-03-16',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
*/
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
vision: true,
|
||||
},
|
||||
contextWindowTokens: 8192,
|
||||
description:
|
||||
'百度自研的旗舰级大规模⼤语⾔模型,覆盖海量中英文语料,具有强大的通用能力,可满足绝大部分对话问答、创作生成、插件应用场景要求;支持自动对接百度搜索插件,保障问答信息时效。',
|
||||
displayName: 'ERNIE 3.5 8K Preview',
|
||||
id: 'ernie-3.5-8k-preview',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.8,
|
||||
output: 2,
|
||||
},
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 128_000,
|
||||
description:
|
||||
'百度自研的旗舰级大规模⼤语⾔模型,覆盖海量中英文语料,具有强大的通用能力,可满足绝大部分对话问答、创作生成、插件应用场景要求;支持自动对接百度搜索插件,保障问答信息时效。',
|
||||
displayName: 'ERNIE 3.5 128K',
|
||||
'文心大模型4.5是百度自主研发的新一代原生多模态基础大模型,通过多个模态联合建模实现协同优化,多模态理解能力优秀;具备更精进的语言能力,理解、生成、逻辑、记忆能力全面提升,去幻觉、逻辑推理、代码能力显著提升。',
|
||||
displayName: 'ERNIE 4.5 8K Preview',
|
||||
enabled: true,
|
||||
id: 'ernie-3.5-128k',
|
||||
id: 'ernie-4.5-8k-preview',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.8,
|
||||
output: 2,
|
||||
input: 4,
|
||||
output: 16,
|
||||
},
|
||||
releasedAt: '2025-03-16',
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
@@ -166,6 +150,66 @@ const wenxinChatModels: AIChatModelCard[] = [
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 8192,
|
||||
description:
|
||||
'百度自研的旗舰级大规模⼤语⾔模型,覆盖海量中英文语料,具有强大的通用能力,可满足绝大部分对话问答、创作生成、插件应用场景要求;支持自动对接百度搜索插件,保障问答信息时效。',
|
||||
displayName: 'ERNIE 3.5 8K',
|
||||
id: 'ernie-3.5-8k',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.8,
|
||||
output: 2,
|
||||
},
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 8192,
|
||||
description:
|
||||
'百度自研的旗舰级大规模⼤语⾔模型,覆盖海量中英文语料,具有强大的通用能力,可满足绝大部分对话问答、创作生成、插件应用场景要求;支持自动对接百度搜索插件,保障问答信息时效。',
|
||||
displayName: 'ERNIE 3.5 8K Preview',
|
||||
id: 'ernie-3.5-8k-preview',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.8,
|
||||
output: 2,
|
||||
},
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
functionCall: true,
|
||||
search: true,
|
||||
},
|
||||
contextWindowTokens: 128_000,
|
||||
description:
|
||||
'百度自研的旗舰级大规模⼤语⾔模型,覆盖海量中英文语料,具有强大的通用能力,可满足绝大部分对话问答、创作生成、插件应用场景要求;支持自动对接百度搜索插件,保障问答信息时效。',
|
||||
displayName: 'ERNIE 3.5 128K',
|
||||
id: 'ernie-3.5-128k',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.8,
|
||||
output: 2,
|
||||
},
|
||||
settings: {
|
||||
searchImpl: 'params',
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
contextWindowTokens: 8192,
|
||||
description:
|
||||
@@ -302,6 +346,150 @@ const wenxinChatModels: AIChatModelCard[] = [
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
reasoning: true,
|
||||
},
|
||||
contextWindowTokens: 32_768,
|
||||
description:
|
||||
'DeepSeek-R1-Distill-Qwen-1.5B是DeepSeek-R1基于Qwen-2.5系列的蒸馏模型。',
|
||||
displayName: 'DeepSeek R1 Distill Qwen 1.5B',
|
||||
id: 'deepseek-r1-distill-qwen-1.5b',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 2,
|
||||
output: 8,
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
reasoning: true,
|
||||
},
|
||||
contextWindowTokens: 32_768,
|
||||
description:
|
||||
'DeepSeek-R1-Distill-Qwen-7B是DeepSeek-R1基于Qwen-2.5系列的蒸馏模型。',
|
||||
displayName: 'DeepSeek R1 Distill Qwen 7B',
|
||||
id: 'deepseek-r1-distill-qwen-7b',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.6,
|
||||
output: 2.4,
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
reasoning: true,
|
||||
},
|
||||
contextWindowTokens: 32_768,
|
||||
description:
|
||||
'DeepSeek-R1-Distill-Qwen-14B是DeepSeek-R1基于Qwen-2.5系列的蒸馏模型。',
|
||||
displayName: 'DeepSeek R1 Distill Qwen 14B',
|
||||
id: 'deepseek-r1-distill-qwen-14b',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 0.6,
|
||||
output: 2.4,
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
reasoning: true,
|
||||
},
|
||||
contextWindowTokens: 32_768,
|
||||
description:
|
||||
'DeepSeek-R1-Distill-Qwen-32B是DeepSeek-R1基于Qwen-2.5系列的蒸馏模型。',
|
||||
displayName: 'DeepSeek R1 Distill Qwen 32B',
|
||||
id: 'deepseek-r1-distill-qwen-32b',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 1.5,
|
||||
output: 6,
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
reasoning: true,
|
||||
},
|
||||
contextWindowTokens: 32_768,
|
||||
description:
|
||||
'DeepSeek-R1-Distill-Llama-8B是DeepSeek-R1基于Llama3.1-8B-Base的蒸馏模型。',
|
||||
displayName: 'DeepSeek R1 Distill Llama 8B',
|
||||
id: 'deepseek-r1-distill-llama-8b',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 1.5,
|
||||
output: 6,
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
reasoning: true,
|
||||
},
|
||||
contextWindowTokens: 32_768,
|
||||
description:
|
||||
'DeepSeek-R1-Distill-Llama-70B是DeepSeek-R1基于Llama3.3-70B-Instruct的蒸馏模型。',
|
||||
displayName: 'DeepSeek R1 Distill Llama 70B',
|
||||
id: 'deepseek-r1-distill-llama-70b',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 2,
|
||||
output: 8,
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
reasoning: true,
|
||||
},
|
||||
contextWindowTokens: 32_768,
|
||||
description:
|
||||
'2025年2月14日首次发布,由千帆大模型研发团队以 Llama3_8B为base模型(Built with Meta Llama)蒸馏所得,蒸馏数据中也同步添加了千帆的语料。',
|
||||
displayName: 'DeepSeek R1 Distill Qianfan Llama 8B',
|
||||
id: 'deepseek-r1-distill-qianfan-llama-8b',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 1,
|
||||
output: 4,
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
reasoning: true,
|
||||
},
|
||||
contextWindowTokens: 32_768,
|
||||
description:
|
||||
'2025年2月14日首次发布,由千帆大模型研发团队以 Llama3_70B为base模型(Built with Meta Llama)蒸馏所得,蒸馏数据中也同步添加了千帆的语料。',
|
||||
displayName: 'DeepSeek R1 Distill Qianfan Llama 70B',
|
||||
id: 'deepseek-r1-distill-qianfan-llama-70b',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 2,
|
||||
output: 8,
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
{
|
||||
abilities: {
|
||||
reasoning: true,
|
||||
},
|
||||
contextWindowTokens: 32_768,
|
||||
description:
|
||||
'通义千问团队推出的高效推理模型,支持消费级硬件部署,具备强大的实时推理能力和与智能体Agent集成的潜力。',
|
||||
displayName: 'QwQ 32B',
|
||||
id: 'qwq-32b',
|
||||
pricing: {
|
||||
currency: 'CNY',
|
||||
input: 2,
|
||||
output: 8,
|
||||
},
|
||||
type: 'chat',
|
||||
},
|
||||
];
|
||||
|
||||
export const allModels = [...wenxinChatModels];
|
||||
|
||||
Reference in New Issue
Block a user