Flowise
Flowise copied to clipboard
[FEATURE] Gemini 1.5 Pro
Describe the feature you'd like
Google just released a public review of their new model, Gemini 1.5 Pro
. It can handle up to 1 million token input capacity. It would really very sweet to test it out before a paid version goes live, on May 2 (I think).
Additional context
Also, they released a new embedding model text-embedding-004
Source: https://developers.googleblog.com/2024/04/gemini-15-pro-in-public-preview-with-new-features.html
[packages/components/nodes/embeddings/GoogleGenerativeAIEmbedding/GoogleGenerativeAIEmbedding.ts]
I would guess the proper code is something like this
// ... existing code ...
async listModels(): Promise<INodeOptionsValue[]> { const models = await getModels(MODEL_TYPE.EMBEDDING, 'googleGenerativeAiEmbeddings') // Add the new model name to the list models.push({ label: 'Gemini 1.5', name: 'gemini-1.5' }) return models }
// ... rest of the code ...
waiitng for the new google/genai version - https://github.com/google/generative-ai-js/pull/96