Llama 4 is here, this is how you can try it!

Meta released their latest model series called LLAMA 4. It's an open-weight's model The herd consists of Llama Scout Llama Maverick Llama Behemoth You can head over to the official release notes here The coolest part about these models are that they go up to 10M context windows, which means that they can read more data, even entire codebases, the Gemini 2.5 Pro that released recently has a 1M context window. You can try this for free on the OpenRouter page. Simply head over to https://openrouter.ai/chat Create your account. Enable this option. Under your Profile => Settings => Privacy. Then when you go to chat add a model The context sizes of this is lower than the original model, but you can wait for that when it releases with the full context size or use OpenRouter with the paid models.

Apr 7, 2025 - 04:10

Llama 4 is here, this is how you can try it!

Meta released their latest model series called LLAMA 4. It's an open-weight's model
The herd consists of

Llama Scout
Llama Maverick
Llama Behemoth

You can head over to the official release notes here

The coolest part about these models are that they go up to 10M context windows, which means that they can read more data, even entire codebases, the Gemini 2.5 Pro that released recently has a 1M context window.

You can try this for free on the OpenRouter page.

Simply head over to https://openrouter.ai/chat Create your account. Enable this option. Under your Profile => Settings => Privacy.

Then when you go to chat add a model

The context sizes of this is lower than the original model, but you can wait for that when it releases with the full context size or use OpenRouter with the paid models.