Skip to content

Commit 0931320

Browse files
sxjeruarvinxx
authored and
bentwnghk
committed
💄 style: Update many models info (lobehub#6205)
* Update stepfun.ts * Update siliconcloud.ts * Update google.ts * Update groq.ts * Update siliconcloud.ts * Update stepfun.ts * Update groq.ts * Update groq.ts * Update hunyuan.ts * Update hunyuan.ts * Update internlm.ts * Update mistral.ts * Update novita.ts * Update sensenova.ts * Update taichu.ts * Update upstage.ts * Update index.test.ts.snap * Update index.test.ts.snap * Update index.test.ts * Update groq.ts --------- Co-authored-by: Arvin Xu <[email protected]>
1 parent 1cc704d commit 0931320

File tree

13 files changed

+608
-197
lines changed

13 files changed

+608
-197
lines changed

‎src/config/aiModels/groq.ts

+111-95
Original file line numberDiff line numberDiff line change
@@ -1,66 +1,62 @@
11
import { AIChatModelCard } from '@/types/aiModel';
22

3+
// https://groq.com/pricing/
4+
// https://console.groq.com/docs/models
5+
36
const groqChatModels: AIChatModelCard[] = [
47
{
58
abilities: {
9+
functionCall: true,
610
reasoning: true,
711
},
8-
contextWindowTokens: 128_000,
9-
description:
10-
'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。',
11-
displayName: 'DeepSeek R1 (Distil-Llama 70B)',
12+
contextWindowTokens: 131_072,
13+
displayName: 'DeepSeek R1 Distill Llama 70B',
1214
enabled: true,
1315
id: 'deepseek-r1-distill-llama-70b',
16+
pricing: {
17+
input: 0.75, // 0.75 - 5.00
18+
output: 0.99, // 0.99 - 5.00
19+
},
1420
type: 'chat',
1521
},
1622
{
17-
abilities: {
18-
functionCall: true,
19-
},
2023
contextWindowTokens: 131_072,
21-
description:
22-
'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。',
23-
displayName: 'Llama 3.3 70B',
24-
enabled: true,
25-
id: 'llama-3.3-70b-versatile',
26-
maxOutput: 8192,
24+
displayName: 'DeepSeek R1 Distill Llama 70B SpecDec',
25+
id: 'deepseek-r1-distill-llama-70b-specdec',
26+
maxOutput: 16_384,
2727
pricing: {
28-
input: 0.05,
29-
output: 0.08,
28+
input: 0.75,
29+
output: 0.99,
3030
},
3131
type: 'chat',
3232
},
3333
{
3434
abilities: {
35-
vision: true,
35+
functionCall: true,
36+
reasoning: true,
3637
},
37-
contextWindowTokens: 8192,
38-
description:
39-
'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。',
40-
displayName: 'Llama 3.2 11B Vision (Preview)',
38+
contextWindowTokens: 131_072,
39+
displayName: 'DeepSeek R1 Distill Qwen 32B',
4140
enabled: true,
42-
id: 'llama-3.2-11b-vision-preview',
43-
maxOutput: 8192,
41+
id: 'deepseek-r1-distill-qwen-32b',
42+
maxOutput: 16_384,
4443
pricing: {
45-
input: 0.05,
46-
output: 0.08,
44+
input: 0.69,
45+
output: 0.69,
4746
},
4847
type: 'chat',
4948
},
5049
{
5150
abilities: {
52-
vision: true,
51+
functionCall: true,
5352
},
5453
contextWindowTokens: 8192,
55-
description:
56-
'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。',
57-
displayName: 'Llama 3.2 90B Vision (Preview)',
58-
enabled: true,
59-
id: 'llama-3.2-90b-vision-preview',
60-
maxOutput: 8192,
54+
description: 'Gemma 2 9B 是一款优化用于特定任务和工具整合的模型。',
55+
displayName: 'Gemma 2 9B',
56+
id: 'gemma2-9b-it',
6157
pricing: {
62-
input: 0.59,
63-
output: 0.79,
58+
input: 0.2,
59+
output: 0.2,
6460
},
6561
type: 'chat',
6662
},
@@ -69,10 +65,8 @@ const groqChatModels: AIChatModelCard[] = [
6965
functionCall: true,
7066
},
7167
contextWindowTokens: 131_072,
72-
description:
73-
'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。',
74-
displayName: 'Llama 3.1 8B',
75-
enabled: true,
68+
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。',
69+
displayName: 'Llama 3.1 8B Instant',
7670
id: 'llama-3.1-8b-instant',
7771
maxOutput: 8192,
7872
pricing: {
@@ -83,103 +77,102 @@ const groqChatModels: AIChatModelCard[] = [
8377
},
8478
{
8579
abilities: {
86-
functionCall: true,
80+
vision: true,
8781
},
8882
contextWindowTokens: 131_072,
89-
description:
90-
'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。',
91-
displayName: 'Llama 3.1 70B',
92-
enabled: true,
93-
id: 'llama-3.1-70b-versatile',
83+
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。',
84+
displayName: 'Llama 3.2 11B Vision (Preview)',
85+
id: 'llama-3.2-11b-vision-preview',
9486
maxOutput: 8192,
9587
pricing: {
96-
input: 0.59,
97-
output: 0.79,
88+
input: 0.18,
89+
output: 0.18,
9890
},
9991
type: 'chat',
10092
},
10193
{
10294
abilities: {
103-
functionCall: true,
95+
vision: true,
10496
},
105-
contextWindowTokens: 8192,
106-
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。',
107-
displayName: 'Llama 3 Groq 8B Tool Use (Preview)',
108-
id: 'llama3-groq-8b-8192-tool-use-preview',
97+
contextWindowTokens: 131_072,
98+
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。',
99+
displayName: 'Llama 3.2 90B Vision (Preview)',
100+
enabled: true,
101+
id: 'llama-3.2-90b-vision-preview',
102+
maxOutput: 8192,
109103
pricing: {
110-
input: 0.19,
111-
output: 0.19,
104+
input: 0.9,
105+
output: 0.9,
112106
},
113107
type: 'chat',
114108
},
115109
{
116-
abilities: {
117-
functionCall: true,
118-
},
119-
contextWindowTokens: 8192,
120-
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。',
121-
displayName: 'Llama 3 Groq 70B Tool Use (Preview)',
122-
id: 'llama3-groq-70b-8192-tool-use-preview',
110+
contextWindowTokens: 131_072,
111+
displayName: 'Llama 3.2 1B (Preview)',
112+
id: 'llama-3.2-1b-preview',
113+
maxOutput: 8192,
123114
pricing: {
124-
input: 0.89,
125-
output: 0.89,
115+
input: 0.04,
116+
output: 0.04,
126117
},
127118
type: 'chat',
128119
},
129120
{
130-
abilities: {
131-
functionCall: true,
121+
contextWindowTokens: 131_072,
122+
displayName: 'Llama 3.2 3B (Preview)',
123+
id: 'llama-3.2-3b-preview',
124+
maxOutput: 8192,
125+
pricing: {
126+
input: 0.06,
127+
output: 0.06,
132128
},
129+
type: 'chat',
130+
},
131+
{
133132
contextWindowTokens: 8192,
134-
description: 'Meta Llama 3 8B 带来优质的推理效能，适合多场景应用需求。',
135-
displayName: 'Meta Llama 3 8B',
136-
id: 'llama3-8b-8192',
133+
displayName: 'Llama 3.3 70B SpecDec',
134+
id: 'llama-3.3-70b-specdec',
137135
pricing: {
138-
input: 0.05,
139-
output: 0.08,
136+
input: 0.59,
137+
output: 0.99,
140138
},
141139
type: 'chat',
142140
},
143141
{
144142
abilities: {
145143
functionCall: true,
146144
},
147-
contextWindowTokens: 8192,
148-
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。',
149-
displayName: 'Meta Llama 3 70B',
150-
id: 'llama3-70b-8192',
145+
contextWindowTokens: 131_072,
146+
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。',
147+
displayName: 'Llama 3.3 70B Versatile',
148+
enabled: true,
149+
id: 'llama-3.3-70b-versatile',
150+
maxOutput: 32_768,
151151
pricing: {
152152
input: 0.59,
153153
output: 0.79,
154154
},
155155
type: 'chat',
156156
},
157157
{
158-
abilities: {
159-
functionCall: true,
160-
},
161158
contextWindowTokens: 8192,
162-
description: 'Gemma 2 9B 是一款优化用于特定任务和工具整合的模型。',
163-
displayName: 'Gemma 2 9B',
164-
enabled: true,
165-
id: 'gemma2-9b-it',
159+
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。',
160+
displayName: 'Llama 3 70B',
161+
id: 'llama3-70b-8192',
166162
pricing: {
167-
input: 0.2,
168-
output: 0.2,
163+
input: 0.59,
164+
output: 0.79,
169165
},
170166
type: 'chat',
171167
},
172168
{
173-
abilities: {
174-
functionCall: true,
175-
},
176169
contextWindowTokens: 8192,
177-
description: 'Gemma 7B 适合中小规模任务处理，兼具成本效益。',
178-
displayName: 'Gemma 7B',
179-
id: 'gemma-7b-it',
170+
description: 'Meta Llama 3 8B 带来优质的推理效能，适合多场景应用需求。',
171+
displayName: 'Llama 3 8B',
172+
id: 'llama3-8b-8192',
180173
pricing: {
181-
input: 0.07,
182-
output: 0.07,
174+
input: 0.05,
175+
output: 0.08,
183176
},
184177
type: 'chat',
185178
},
@@ -189,7 +182,7 @@ const groqChatModels: AIChatModelCard[] = [
189182
},
190183
contextWindowTokens: 32_768,
191184
description: 'Mixtral 8x7B 提供高容错的并行计算能力，适合复杂任务。',
192-
displayName: 'Mixtral 8x7B',
185+
displayName: 'Mixtral 8x7B Instruct',
193186
id: 'mixtral-8x7b-32768',
194187
pricing: {
195188
input: 0.24,
@@ -199,12 +192,35 @@ const groqChatModels: AIChatModelCard[] = [
199192
},
200193
{
201194
abilities: {
202-
vision: true,
195+
functionCall: true,
196+
},
197+
contextWindowTokens: 131_072,
198+
displayName: 'Qwen 2.5 32B',
199+
id: 'qwen-2.5-32b',
200+
pricing: {
201+
input: 0.79,
202+
output: 0.79,
203+
},
204+
type: 'chat',
205+
},
206+
{
207+
contextWindowTokens: 131_072,
208+
displayName: 'Qwen 2.5 Coder 32B',
209+
id: 'qwen-2.5-coder-32b',
210+
pricing: {
211+
input: 0.79,
212+
output: 0.79,
213+
},
214+
type: 'chat',
215+
},
216+
{
217+
contextWindowTokens: 8192,
218+
displayName: 'Llama Guard 3 8B',
219+
id: 'llama-guard-3-8b',
220+
pricing: {
221+
input: 0.2,
222+
output: 0.2,
203223
},
204-
contextWindowTokens: 4096,
205-
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。',
206-
displayName: 'LLaVA 1.5 7B',
207-
id: 'llava-v1.5-7b-4096-preview',
208224
type: 'chat',
209225
},
210226
];

‎src/config/aiModels/hunyuan.ts

+36-4
Original file line numberDiff line numberDiff line change
@@ -1,5 +1,7 @@
11
import { AIChatModelCard } from '@/types/aiModel';
22

3+
// https://cloud.tencent.com/document/product/1729/104753
4+
35
const hunyuanChatModels: AIChatModelCard[] = [
46
{
57
contextWindowTokens: 256_000,
@@ -30,7 +32,7 @@ const hunyuanChatModels: AIChatModelCard[] = [
3032
input: 0.8,
3133
output: 2,
3234
},
33-
releasedAt: '2024-10-28',
35+
releasedAt: '2025-02-10',
3436
type: 'chat',
3537
},
3638
{
@@ -46,7 +48,7 @@ const hunyuanChatModels: AIChatModelCard[] = [
4648
input: 0.5,
4749
output: 2,
4850
},
49-
releasedAt: '2024-10-28',
51+
releasedAt: '2025-02-10',
5052
type: 'chat',
5153
},
5254
{
@@ -56,7 +58,7 @@ const hunyuanChatModels: AIChatModelCard[] = [
5658
contextWindowTokens: 32_000,
5759
description:
5860
'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。',
59-
displayName: 'Hunyuan Turbo',
61+
displayName: 'Hunyuan Turbo Latest',
6062
enabled: true,
6163
id: 'hunyuan-turbo-latest',
6264
maxOutput: 4000,
@@ -135,7 +137,7 @@ const hunyuanChatModels: AIChatModelCard[] = [
135137
input: 4,
136138
output: 12,
137139
},
138-
releasedAt: '2024-11-20',
140+
releasedAt: '2025-02-10',
139141
type: 'chat',
140142
},
141143
{
@@ -264,6 +266,36 @@ const hunyuanChatModels: AIChatModelCard[] = [
264266
releasedAt: '2024-07-04',
265267
type: 'chat',
266268
},
269+
{
270+
contextWindowTokens: 8000,
271+
description:
272+
'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。',
273+
displayName: 'Hunyuan Translation',
274+
id: 'hunyuan-translation',
275+
maxOutput: 4000,
276+
pricing: {
277+
currency: 'CNY',
278+
input: 25,
279+
output: 75,
280+
},
281+
releasedAt: '2024-10-25',
282+
type: 'chat',
283+
},
284+
{
285+
contextWindowTokens: 8000,
286+
description:
287+
'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。',
288+
displayName: 'Hunyuan Translation Lite',
289+
id: 'hunyuan-translation-lite',
290+
maxOutput: 4000,
291+
pricing: {
292+
currency: 'CNY',
293+
input: 5,
294+
output: 15,
295+
},
296+
releasedAt: '2024-11-25',
297+
type: 'chat',
298+
},
267299
];
268300

269301
export const allModels = [...hunyuanChatModels];

0 commit comments

Comments
 (0)