From 68e4379e637df5c2031d9c04280862187e36a550 Mon Sep 17 00:00:00 2001
From: Caijinglong <cjl_spy@163.com>
Date: Tue, 10 Dec 2024 10:36:12 +0800
Subject: [PATCH] =?UTF-8?q?=F0=9F=92=84=20style:=20Update=20groq,=20add=20?=
 =?UTF-8?q?llama3.3=20(#4942)?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

- Add llama3.3
- Set `modelList.showModelFetcher` to true
---
 src/config/modelProviders/groq.ts | 15 +++++++++++++++
 1 file changed, 15 insertions(+)

diff --git a/src/config/modelProviders/groq.ts b/src/config/modelProviders/groq.ts
index 33b847d495e2c..2c153008f4c58 100644
--- a/src/config/modelProviders/groq.ts
+++ b/src/config/modelProviders/groq.ts
@@ -4,6 +4,20 @@ import { ModelProviderCard } from '@/types/llm';
 const Groq: ModelProviderCard = {
   chatModels: [
     // TODO: During preview launch, Groq is limiting 3.2 models to max_tokens of 8k.
+    {
+      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,
+      functionCall: true,
+      id: 'llama-3.3-70b-versatile',
+      maxOutput: 8192,
+      pricing: {
+        input: 0.05,
+        output: 0.08,
+      },
+      tokens: 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。',
@@ -159,6 +173,7 @@ const Groq: ModelProviderCard = {
   description:
     'Groq 的 LPU 推理引擎在最新的独立大语言模型（LLM）基准测试中表现卓越，以其惊人的速度和效率重新定义了 AI 解决方案的标准。Groq 是一种即时推理速度的代表，在基于云的部署中展现了良好的性能。',
   id: 'groq',
+  modelList: { showModelFetcher: true },
   modelsUrl: 'https://console.groq.com/docs/models',
   name: 'Groq',
   proxyUrl: {