Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

💄 style: Update many models info #6205

Merged
merged 22 commits into from
Feb 23, 2025
Merged
Show file tree
Hide file tree
Changes from all commits
Commits
File filter

Filter by extension

Filter by extension

Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
206 changes: 111 additions & 95 deletions src/config/aiModels/groq.ts
Original file line number Diff line number Diff line change
@@ -1,66 +1,62 @@
import { AIChatModelCard } from '@/types/aiModel';

// https://groq.com/pricing/
// https://console.groq.com/docs/models

const groqChatModels: AIChatModelCard[] = [
{
abilities: {
functionCall: true,
reasoning: true,
},
contextWindowTokens: 128_000,
description:
'DeepSeek R1——DeepSeek 套件中更大更智能的模型——被蒸馏到 Llama 70B 架构中。基于基准测试和人工评估,该模型比原始 Llama 70B 更智能,尤其在需要数学和事实精确性的任务上表现出色。',
displayName: 'DeepSeek R1 (Distil-Llama 70B)',
contextWindowTokens: 131_072,
displayName: 'DeepSeek R1 Distill Llama 70B',
enabled: true,
id: 'deepseek-r1-distill-llama-70b',
pricing: {
input: 0.75, // 0.75 - 5.00
output: 0.99, // 0.99 - 5.00
},
type: 'chat',
},
{
abilities: {
functionCall: true,
},
contextWindowTokens: 131_072,
description:
'Meta Llama 3.3 多语言大语言模型 ( LLM ) 是 70B(文本输入/文本输出)中的预训练和指令调整生成模型。 Llama 3.3 指令调整的纯文本模型针对多语言对话用例进行了优化,并且在常见行业基准上优于许多可用的开源和封闭式聊天模型。',
displayName: 'Llama 3.3 70B',
enabled: true,
id: 'llama-3.3-70b-versatile',
maxOutput: 8192,
displayName: 'DeepSeek R1 Distill Llama 70B SpecDec',
id: 'deepseek-r1-distill-llama-70b-specdec',
maxOutput: 16_384,
pricing: {
input: 0.05,
output: 0.08,
input: 0.75,
output: 0.99,
},
type: 'chat',
},
{
abilities: {
vision: true,
functionCall: true,
reasoning: true,
},
contextWindowTokens: 8192,
description:
'Llama 3.2 旨在处理结合视觉和文本数据的任务。它在图像描述和视觉问答等任务中表现出色,跨越了语言生成和视觉推理之间的鸿沟。',
displayName: 'Llama 3.2 11B Vision (Preview)',
contextWindowTokens: 131_072,
displayName: 'DeepSeek R1 Distill Qwen 32B',
enabled: true,
id: 'llama-3.2-11b-vision-preview',
maxOutput: 8192,
id: 'deepseek-r1-distill-qwen-32b',
maxOutput: 16_384,
pricing: {
input: 0.05,
output: 0.08,
input: 0.69,
output: 0.69,
},
type: 'chat',
},
{
abilities: {
vision: true,
functionCall: true,
},
contextWindowTokens: 8192,
description:
'Llama 3.2 旨在处理结合视觉和文本数据的任务。它在图像描述和视觉问答等任务中表现出色,跨越了语言生成和视觉推理之间的鸿沟。',
displayName: 'Llama 3.2 90B Vision (Preview)',
enabled: true,
id: 'llama-3.2-90b-vision-preview',
maxOutput: 8192,
description: 'Gemma 2 9B 是一款优化用于特定任务和工具整合的模型。',
displayName: 'Gemma 2 9B',
id: 'gemma2-9b-it',
pricing: {
input: 0.59,
output: 0.79,
input: 0.2,
output: 0.2,
},
type: 'chat',
},
Expand All @@ -69,10 +65,8 @@ const groqChatModels: AIChatModelCard[] = [
functionCall: true,
},
contextWindowTokens: 131_072,
description:
'Llama 3.1 8B 是一款高效能模型,提供了快速的文本生成能力,非常适合需要大规模效率和成本效益的应用场景。',
displayName: 'Llama 3.1 8B',
enabled: true,
description: 'Llama 3.1 8B 是一款高效能模型,提供了快速的文本生成能力,非常适合需要大规模效率和成本效益的应用场景。',
displayName: 'Llama 3.1 8B Instant',
id: 'llama-3.1-8b-instant',
maxOutput: 8192,
pricing: {
Expand All @@ -83,103 +77,102 @@ const groqChatModels: AIChatModelCard[] = [
},
{
abilities: {
functionCall: true,
vision: true,
},
contextWindowTokens: 131_072,
description:
'Llama 3.1 70B 提供更强大的AI推理能力,适合复杂应用,支持超多的计算处理并保证高效和准确率。',
displayName: 'Llama 3.1 70B',
enabled: true,
id: 'llama-3.1-70b-versatile',
description: 'Llama 3.2 旨在处理结合视觉和文本数据的任务。它在图像描述和视觉问答等任务中表现出色,跨越了语言生成和视觉推理之间的鸿沟。',
displayName: 'Llama 3.2 11B Vision (Preview)',
id: 'llama-3.2-11b-vision-preview',
maxOutput: 8192,
pricing: {
input: 0.59,
output: 0.79,
input: 0.18,
output: 0.18,
},
type: 'chat',
},
{
abilities: {
functionCall: true,
vision: true,
},
contextWindowTokens: 8192,
description: 'Llama 3 Groq 8B Tool Use 是针对高效工具使用优化的模型,支持快速并行计算。',
displayName: 'Llama 3 Groq 8B Tool Use (Preview)',
id: 'llama3-groq-8b-8192-tool-use-preview',
contextWindowTokens: 131_072,
description: 'Llama 3.2 旨在处理结合视觉和文本数据的任务。它在图像描述和视觉问答等任务中表现出色,跨越了语言生成和视觉推理之间的鸿沟。',
displayName: 'Llama 3.2 90B Vision (Preview)',
enabled: true,
id: 'llama-3.2-90b-vision-preview',
maxOutput: 8192,
pricing: {
input: 0.19,
output: 0.19,
input: 0.9,
output: 0.9,
},
type: 'chat',
},
{
abilities: {
functionCall: true,
},
contextWindowTokens: 8192,
description: 'Llama 3 Groq 70B Tool Use 提供强大的工具调用能力,支持复杂任务的高效处理。',
displayName: 'Llama 3 Groq 70B Tool Use (Preview)',
id: 'llama3-groq-70b-8192-tool-use-preview',
contextWindowTokens: 131_072,
displayName: 'Llama 3.2 1B (Preview)',
id: 'llama-3.2-1b-preview',
maxOutput: 8192,
pricing: {
input: 0.89,
output: 0.89,
input: 0.04,
output: 0.04,
},
type: 'chat',
},
{
abilities: {
functionCall: true,
contextWindowTokens: 131_072,
displayName: 'Llama 3.2 3B (Preview)',
id: 'llama-3.2-3b-preview',
maxOutput: 8192,
pricing: {
input: 0.06,
output: 0.06,
},
type: 'chat',
},
{
contextWindowTokens: 8192,
description: 'Meta Llama 3 8B 带来优质的推理效能,适合多场景应用需求。',
displayName: 'Meta Llama 3 8B',
id: 'llama3-8b-8192',
displayName: 'Llama 3.3 70B SpecDec',
id: 'llama-3.3-70b-specdec',
pricing: {
input: 0.05,
output: 0.08,
input: 0.59,
output: 0.99,
},
type: 'chat',
},
{
abilities: {
functionCall: true,
},
contextWindowTokens: 8192,
description: 'Meta Llama 3 70B 提供无与伦比的复杂性处理能力,为高要求项目量身定制。',
displayName: 'Meta Llama 3 70B',
id: 'llama3-70b-8192',
contextWindowTokens: 131_072,
description: 'Meta Llama 3.3 多语言大语言模型 ( LLM ) 是 70B(文本输入/文本输出)中的预训练和指令调整生成模型。 Llama 3.3 指令调整的纯文本模型针对多语言对话用例进行了优化,并且在常见行业基准上优于许多可用的开源和封闭式聊天模型。',
displayName: 'Llama 3.3 70B Versatile',
enabled: true,
id: 'llama-3.3-70b-versatile',
maxOutput: 32_768,
pricing: {
input: 0.59,
output: 0.79,
},
type: 'chat',
},
{
abilities: {
functionCall: true,
},
contextWindowTokens: 8192,
description: 'Gemma 2 9B 是一款优化用于特定任务和工具整合的模型。',
displayName: 'Gemma 2 9B',
enabled: true,
id: 'gemma2-9b-it',
description: 'Meta Llama 3 70B 提供无与伦比的复杂性处理能力,为高要求项目量身定制。',
displayName: 'Llama 3 70B',
id: 'llama3-70b-8192',
pricing: {
input: 0.2,
output: 0.2,
input: 0.59,
output: 0.79,
},
type: 'chat',
},
{
abilities: {
functionCall: true,
},
contextWindowTokens: 8192,
description: 'Gemma 7B 适合中小规模任务处理,兼具成本效益。',
displayName: 'Gemma 7B',
id: 'gemma-7b-it',
description: 'Meta Llama 3 8B 带来优质的推理效能,适合多场景应用需求。',
displayName: 'Llama 3 8B',
id: 'llama3-8b-8192',
pricing: {
input: 0.07,
output: 0.07,
input: 0.05,
output: 0.08,
},
type: 'chat',
},
Expand All @@ -189,7 +182,7 @@ const groqChatModels: AIChatModelCard[] = [
},
contextWindowTokens: 32_768,
description: 'Mixtral 8x7B 提供高容错的并行计算能力,适合复杂任务。',
displayName: 'Mixtral 8x7B',
displayName: 'Mixtral 8x7B Instruct',
id: 'mixtral-8x7b-32768',
pricing: {
input: 0.24,
Expand All @@ -199,12 +192,35 @@ const groqChatModels: AIChatModelCard[] = [
},
{
abilities: {
vision: true,
functionCall: true,
},
contextWindowTokens: 131_072,
displayName: 'Qwen 2.5 32B',
id: 'qwen-2.5-32b',
pricing: {
input: 0.79,
output: 0.79,
},
type: 'chat',
},
{
contextWindowTokens: 131_072,
displayName: 'Qwen 2.5 Coder 32B',
id: 'qwen-2.5-coder-32b',
pricing: {
input: 0.79,
output: 0.79,
},
type: 'chat',
},
{
contextWindowTokens: 8192,
displayName: 'Llama Guard 3 8B',
id: 'llama-guard-3-8b',
pricing: {
input: 0.2,
output: 0.2,
},
contextWindowTokens: 4096,
description: 'LLaVA 1.5 7B 提供视觉处理能力融合,通过视觉信息输入生成复杂输出。',
displayName: 'LLaVA 1.5 7B',
id: 'llava-v1.5-7b-4096-preview',
type: 'chat',
},
];
Expand Down
40 changes: 36 additions & 4 deletions src/config/aiModels/hunyuan.ts
Original file line number Diff line number Diff line change
@@ -1,5 +1,7 @@
import { AIChatModelCard } from '@/types/aiModel';

// https://cloud.tencent.com/document/product/1729/104753

const hunyuanChatModels: AIChatModelCard[] = [
{
contextWindowTokens: 256_000,
Expand Down Expand Up @@ -30,7 +32,7 @@ const hunyuanChatModels: AIChatModelCard[] = [
input: 0.8,
output: 2,
},
releasedAt: '2024-10-28',
releasedAt: '2025-02-10',
type: 'chat',
},
{
Expand All @@ -46,7 +48,7 @@ const hunyuanChatModels: AIChatModelCard[] = [
input: 0.5,
output: 2,
},
releasedAt: '2024-10-28',
releasedAt: '2025-02-10',
type: 'chat',
},
{
Expand All @@ -56,7 +58,7 @@ const hunyuanChatModels: AIChatModelCard[] = [
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',
displayName: 'Hunyuan Turbo Latest',
enabled: true,
id: 'hunyuan-turbo-latest',
maxOutput: 4000,
Expand Down Expand Up @@ -135,7 +137,7 @@ const hunyuanChatModels: AIChatModelCard[] = [
input: 4,
output: 12,
},
releasedAt: '2024-11-20',
releasedAt: '2025-02-10',
type: 'chat',
},
{
Expand Down Expand Up @@ -264,6 +266,36 @@ 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:
'混元翻译模型支持自然语言对话式翻译;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。',
displayName: 'Hunyuan Translation Lite',
id: 'hunyuan-translation-lite',
maxOutput: 4000,
pricing: {
currency: 'CNY',
input: 5,
output: 15,
},
releasedAt: '2024-11-25',
type: 'chat',
},
];

export const allModels = [...hunyuanChatModels];
Expand Down
Loading