README.md

---
title: Space Multiverse
emoji: 🤖
colorFrom: yellow
colorTo: blue
sdk: docker
app_port: 7860
short_description: ML Intern V2 (FREE & SAVE MODEL)
hf_oauth: true
hf_oauth_expiration_minutes: 43200
hf_oauth_scopes:
- read-repos
- write-repos
- contribute-repos
- manage-repos
- write-collections
- inference-api
- jobs
- write-discussions
tags:
- ml-intern
---

# Space Multiverse — ML Intern V2 (FREE & SAVE MODEL)

<!-- rebuild: add-deepseek-v4-flash-latest-2026-08-02T09:00:00Z -->

- Added **DeepSeek V4 Flash latest** via OpenRouter (path: `openai/deepseek/deepseek-v4-flash-latest`)
- Replaced GPT-5.5 model with **Laguna S 2.1:free** (path: `openai/poolside/laguna-s-2.1:free`)
- Added OpenRouter routing (catch-all `openai/` prefix → OpenRouter API)
- Default: `openai/tencent/hy3:free`
- Models: Tencent HY3:free, Gemma 4 31B:free, Llama 3.3 70B:free, **Laguna M.1:free**, **Laguna S 2.1:free**, Llama 3.1 8B, DeepSeek V4 Flash latest
- Ollama: không thể dùng cho HF Space (local inference, không có cloud API free cho use case này)