| import Groq from "groq-sdk"; |
|
|
| export async function callGroq({ |
| model, messages, temperature = 1.0, maxTokens = 2048, |
| reasoning = false, stream = false, apiKey, onToken, |
| }) { |
| const client = new Groq({ apiKey }); |
| const modelName = model.replace(/^groq:/, ""); |
|
|
| const chatMessages = messages.map((m) => ({ role: m.role, content: m.content })); |
|
|
| const params = { |
| model: modelName, |
| messages: chatMessages, |
| temperature: Number(temperature), |
| max_tokens: Number(maxTokens), |
| stream, |
| }; |
|
|
| if (stream) { |
| const completion = await client.chat.completions.create({ ...params, stream: true }); |
| let fullText = ""; |
| for await (const chunk of completion) { |
| const delta = chunk.choices?.[0]?.delta?.content; |
| if (delta) { fullText += delta; onToken?.(delta); } |
| } |
| return { text: fullText, usage: null }; |
| } |
|
|
| const completion = await client.chat.completions.create(params); |
| return { |
| text: completion.choices?.[0]?.message?.content || "", |
| usage: { |
| promptTokens: completion.usage?.prompt_tokens, |
| completionTokens: completion.usage?.completion_tokens, |
| }, |
| }; |
| } |
|
|
