Meta AI Models on XALEN

Meta open-source AI. All Llama models on XALEN.

Founded: 2013 · HQ: Menlo Park, USA · 6 models available

Available Models

Llama 4 Scout

Open Source

Meta latest MoE model with 512K context.

Params: 109B (17B active) Context: 512K Input: $0.05 Output: $0.08 Latency: ~350ms

Llama 4 Maverick

Open Source

Meta largest MoE model with superior reasoning.

Params: 400B (17B active) Context: 256K Input: $0.07 Output: $0.12 Latency: ~450ms

Llama 3.3 70B

Open Source

Proven 70B instruction model with strong Hindi/Tamil.

Params: 70B Context: 128K Input: $0.04 Output: $0.06 Latency: ~300ms

Llama 3.1 405B

Open Source

Largest open model for maximum quality.

Params: 405B Context: 128K Input: $0.08 Output: $0.14 Latency: ~600ms

Llama 3.1 8B Turbo

Compact

Ultra-fast compact model for high-throughput.

Params: 8B Context: 128K Input: $0.01 Output: $0.02 Latency: ~60ms

Llama 3.2 90B Vision

Vision

Vision-language model for analyzing images with text understanding.

Params: 90B Context: 128K Input: $0.06 Output: $0.10 Latency: ~500ms

Quick Start

import XALEN from 'xalen-sdk';

const client = new XALEN({ apiKey: 'xln_test_YOUR_KEY' });

const response = await client.chat.completions.create({
  model: 'llama-4-scout',
  messages: [{ role: 'user', content: 'Hello!' }]
});

Start Using Meta AI Models

All models available through a single API. No separate accounts needed.

Get API Key

Last updated: 2026-05-21.