Multimodal

Llama license / check exact model cardOpen weightsUpdated August 2026Frontier 2026

Llama 4 Scout

Open-weight Llama 4 model positioned for multimodal and long-context workflows.

Meta · Llama

Editorial review

Reviewed byOpenSourcesAI EditorialLast updatedAugust 2026SourcesOpenSourcesAI checker catalog + Hugging Face model card (meta-llama/Llama-4-Scout-17B-16E-Instruct)

Model checkpoints, context windows, provider support, local runtime compatibility, and license terms can change quickly. Verify the exact model card before production or commercial use.

Best for

Teams evaluating Llama-family models for multimodal assistant, long-context, and application workflows.

Who should use it

  • Teams evaluating Llama-family models for multimodal assistant, long-context, and application workflows.
  • Builders who want local or self-hosted testing options.

Common workflows

  • Multimodal and long-context assistant workflows
  • multimodal workflows
  • long-context workflows
  • open weights workflows
  • llama workflows

Deployment and hardware notes

Likely best evaluated through compatible hosted or server-class runtimes unless smaller or quantized builds fit your hardware.

License and usage notes

Llama license / check exact model card. Open weights. Verify the exact model card and license terms for the checkpoint or hosted provider you use.

Strengths

  • Open weights model option for Llama workflows.
  • Teams evaluating Llama-family models for multimodal assistant, long-context, and application workflows.
  • Evaluate local fit with the exact checkpoint and quantization available for your runtime.
  • Tracked as Frontier 2026 in the OpenSourcesAI model directory.

Limitations

  • Review the exact model card, license, runtime support, and serving requirements before production use.
  • Likely best evaluated through compatible hosted or server-class runtimes unless smaller or quantized builds fit your hardware.
  • Context window and limits: 10,485,760 tokens.
  • Verify the exact model card, provider docs, license, and serving support before production use.

Local workflow notes

Evaluate local fit with the exact checkpoint and quantization available for your runtime.

Local runtimes: Transformers, vLLM where supported

Platforms: Windows, macOS, Linux, Self-hosted servers

Will Llama 4 Scout run on your machine?

Llama 4 Scout is 109B parameters and needs 64.5 GB of VRAM at Q4_K_M — 63 GB of weights plus 1.5 GB of runtime overhead for the inference server itself.

VRAM by quantization

QuantizationWeightsNeeds (with overhead)Quality
Q4_K_M63 GB64.5 GBgood

Fit on common hardware at Q4_K_M

HardwareMemory the model can useSystem RAMVerdict
CPU OnlyNone (CPU only)16 GBToo large
RTX 4060 Laptop8 GB16 GBToo large
RTX 3060 (12GB)12 GB32 GBToo large
RTX 4060 Ti (16GB)16 GB32 GBToo large
RTX 309024 GB64 GBToo large
Apple Silicon (Unified Memory) 36 GB27 GB of 36 GB36 GBToo large
RTX 509032 GB64 GBToo large
Apple Silicon (Unified Memory) 48 GB36 GB of 48 GB48 GBToo large
Apple Silicon (Unified Memory) 64 GB48 GB of 64 GB64 GBToo large
Apple Silicon (Unified Memory) 96 GB72 GB of 96 GB96 GBComfortable
Apple Silicon (Unified Memory) 128 GB96 GB of 128 GB128 GBComfortable
Apple Silicon (Unified Memory) 192 GB144 GB of 192 GB192 GBComfortable

Comfortable means VRAM clears the requirement by 2 GB or more. Tight means it covers the requirement with no margin. CPU offload means the model does not fit in VRAM but system RAM is at least 1.6× the weights, so it will run at reduced speed — expect roughly 1–5 tokens per second. Figures are weights plus a fixed runtime overhead and exclude KV-cache growth, which scales with context length.

Apple Silicon shares one pool of memory between the system and the GPU, so a model cannot use all of it. These rows apply the same 75% usable fraction the Compatibility Checker uses, which is why a 36 GB Mac is graded on less than 36 GB.

Check Llama 4 Scout against your exact hardwareHow these verdicts are calculated

VRAM fit by quantization level

Enter your GPU VRAM below to see which quantization of Llama 4 Scout fits and get the Ollama run command.

Frontier-model verification note

This page is written to stay accurate as of the latest available 2026 public model information. Availability, licenses, context windows, API support, pricing, benchmark standing, and local-serving support can change quickly. Verify the official model card, provider docs, and license before using this model in production or commercial workflows.

Sources to verify

OpenSourcesAI checker catalog + Hugging Face model card (meta-llama/Llama-4-Scout-17B-16E-Instruct)

Related resources

Continue with model source notes, local tools, and implementation guides related to this model.

Llama family hubAI tools directoryModel source notesModels directoryGuides library
Hardware~63 GB at Q4_K_M (109B total parameters, MoE) — multi-GPU classRuntimevLLM, Transformers, hosted providers where supportedContext10,485,760 tokensLast updated2026
OpenSourcesAI checker catalog + Hugging Face model card (meta-llama/Llama-4-Scout-17B-16E-Instruct) →

Model ecosystem connections

Use these next-step links to move from this profile into related tools, comparisons, guides, stacks, and curated shortlists.

Recommended runtimes and tools

Transformers →vLLM →Ollama →LM Studio →llama.cpp →

Setup and deployment

What is quantization? GGUF, Q4, Q8 explained →RunPod — managed GPU cloud →Lambda Labs — A100 / H100 on demand →Vast.ai — crowdsourced GPU marketplace →

Related model pages

Llama family hub →Llama Guard 3 →Llama 3.1 405B Instruct →Llama 3.1 70B Instruct →Phi-4 Multimodal →Llama 4 Maverick →Llama 3.3 70B →

Guides, stacks, and comparisons

AI tools directory →Model selection guide →All model comparisons →

Related models

Llama Guard 3Safety checks, guardrail experiments, and policy classification workflows.Llama 3.1 405B InstructServer-class assistant evaluation and comparisons against smaller Llama variants.Llama 3.1 70B InstructTeams comparing widely supported Llama-family 70B-class models.Phi-4 MultimodalSmall multimodal and edge experimentsLlama 4 MaverickBuilders comparing current Llama-family models for assistant, multimodal, and reasoning-oriented workflows.Llama 3.3 70B70B open-weight model in the Llama family.

Ready to run this model locally?

Find a compatible interface in our Local AI Tools directory →