Veliki jezični modeli (LLM-ovi) su napredni AI sustavi obučeni za generiranje odgovora sličnih ljudskim. Braveov AI Asistent, Leo, nudi opcije za korisnike kako bi odabrali odgovarajući model i konfiguraciju modela prema njihovim potrebama privatnosti i funkcionalnosti. These LLMs include the latest Qwen, Meta's Llama, and Anthropic's Claude Haiku and Claude Sonnet models, all of which Brave hosts through our own secure infrastructure. Iako modeli imaju različite sposobnosti, Brave osigurava privatnost korisnika bez obzira na to koji se model koristi i kako se poslužuje.

LLM-ovi Objašnjeni

Larger models like Qwen and Llama have more parameters and complexity, allowing them to handle more complex tasks but also requiring more computing power to run efficiently. Models trained on diverse data like Claude Sonnet have broad knowledge, while specialized models like Claude Haiku have domain expertise like coding or legal analysis. Neki LLM-ovi su izvrsni u generiranju teksta, dok su drugi bolji u odgovaranju na pitanja ili zadacima analize podataka. Brzina će ovisiti o veličini modela, snazi infrastrukture koja pokreće model, tehnikama korištenim za optimizaciju, količini sadržaja koju obrađuje i složenosti zahtjeva.

Importantly, all models in Brave Leo are managed and operated exclusively by Brave in secure environments, with no data sharing between Brave and the original model creators. For example, your use of Llama in Brave Leo is totally independent from the model creator, Meta. 

Brave Leo nudi opcije za odabir odgovarajućeg modela prema vašim zahtjevima. You can use the Brave-hosted Automatic mode, manually select models from the list provided, or, for advanced users, connect Leo to your own model (See BYOM).

Privacy Protections

Models will change from time-to-time, but Brave's privacy protections will protect you no matter what model you are using:

• Vaši razgovori se ne pohranjuju niti logiraju od strane Brave-a nakon što se generira odgovor. No personal data such as your IP address or conversation transcripts are retained. A vaši razgovori se ne koriste za potrebe obuke modela.
• Podaci o razgovoru se brišu iz Braveovih servera odmah nakon završetka sjednice chata.
• Your chat history exists only on your device in your local storage. Chats can be removed from your device in Settings by selection the option to “Clear Leo data”.
• Chats can be temporary, meaning they aren’t kept in your local storage, and chat history can be disabled entirely so conversations don’t exist on your device or anywhere else after you close them.

Odabir modela

The Leo panel includes a 3-dot icon in the header where you can change the Settings. You will see options for Language Models which you can select manually or choose to continue with the default `Automatic mode`.

You can also open advanced settings which will take you to brave://settings/leo-ai where you can select which model you want to use by default every time you use Leo.

Modeli i značajke dostupni u Brave Leo ovise o tome koristite li besplatnu verziju ili imate premium pretplatu: