Meta AI
Meta · General AI
Meta's assistant on the Llama model family, embedded in Meta's apps and standalone. Convenient where you already message; personalization is lighter than dedicated assistants.
Best for: Casual assistant inside WhatsApp, Instagram and Messenger.
| Models | Llama model family |
|---|---|
| Free plan | Yes |
| Paid plans | Primarily free (offerings vary by region) |
| API | Llama models are openly available to developers |
| Custom instructions | No dedicated persistent instruction box on most surfaces. Paste per conversation |
| Memory | Memory features on some surfaces/regions |
| File upload | Varies by surface |
| Web search | Yes |
| Context window | Model-dependent |
| Apps | WhatsApp · Instagram · Messenger · Web |
Facts last reviewed 2026-08-10, capabilities and plans change frequently; verify on the official site before relying on a detail.
Models, limits and pricing
| Current models | Llama 4 Maverick, flagship open-weight multimodal MoE Llama 4 Scout, single-H100 open-weight MoE with 10M context Llama 4 Behemoth, teacher model, not released Meta does not disclose which build serves the consumer assistant |
|---|---|
| Context window | 10M tokens (Llama 4 Scout); 1M tokens (Llama 4 Maverick) |
| Parameters | Llama 4 Scout 109B total / 17B active; Maverick 400B total / 17B active |
| Modalities | Text + image in; text out; image and video generation; voice in Meta apps |
| Consumer pricing |
|
| API pricing | See vendor pricing page (Meta Model API) |
Pricing and model line-ups verified 2026-08-11 from source 1 ↗ · source 2 ↗ · source 3 ↗. These change often. The vendor page is authoritative.
Notable
- Reach. It's inside the messaging apps
- Open-weight model family behind it