Llama 4 Scout

At a glance

What Llama 4 Scout does

Vision, Tool use, MoE, Function calling, 10M context, Single-GPU friendly

Best for

On-prem 10M-token context analysis, doc/codebase RAG without external chunking

Works well with

Conflicts & caveats

• Trae's built-in model menu may not include Llama 4 Scout — Trae primarily ships with Claude/GPT/Gemini/DeepSeek.
• Google Antigravity is built around Gemini 3 Pro; Claude and recent GPT-5 models also work, but Llama 4 Scout is not in the supported model list.
• Kiro's spec mode is tuned for Claude models — Llama 4 Scout may work but spec-driven flows are best with Claude.
• Cursor's built-in model menu may not include Llama 4 Scout — verify it's available in your Cursor model picker.
• OpenAI Codex (VS Code) is tied to OpenAI's GPT-5.1 Codex family — Llama 4 Scout is not selectable inside this integration.
• Gemini CLI is locked to Google Gemini models — your LLM is Llama 4 Scout.
• Privacy conflict: Self-hosted Llama 4 Scout sends code to cloud @codebase (Cursor). Use local context (Continue indexing, ChromaDB, LanceDB, pgvector, Vespa self-hosted) for true privacy.
• SWE-agent with Llama 4 Scout (≤8K max output) may struggle with multi-file edits. Use models with 16K+ output capacity.
• Claude Code only works with Claude models — your LLM is Llama 4 Scout.
• Devin uses Cognition's proprietary model internally — your selected Llama 4 Scout will be ignored.
• OpenAI Codex (Cloud) runs GPT-5.1 Codex internally — your selected Llama 4 Scout will be ignored.
• Jules runs Gemini models internally — your selected Llama 4 Scout will be ignored.
• The Claude Code SDK is tuned for Claude models — Llama 4 Scout may work via the model adapter but loses Claude-specific features (thinking, prompt caching, sub-agents).
• Lovable uses a fixed internal model — your selected Llama 4 Scout will be ignored.
• Replit Agent 3 uses its own model selection internally — your selected Llama 4 Scout will be ignored.