Llama 4 Maverick 400B
Large multimodal MoE model intended for multi-GPU servers and very high-memory systems.
DeviceTerra has not lab-tested this exact artifact on every hardware combination. RAM and VRAM floors are LocalLens planning estimates, not publisher guarantees.
Will it run?
LocalLens uses 320 GB as a conservative planning floor for this profile and its listed quantization. This is not measured speed. Processor, backend support, context length, memory bandwidth and other applications still affect the result.
Check your complete system →What is it useful for?
Suggested Ollama artifact
Confirm the tag on the official page before downloading, then run:
ollama run llama4:maverickStart with a representative non-sensitive task. Record the exact runtime, quantization, context, load time, generation speed, memory use and answer quality.
Evidence and limits
Published context: Very long context; practical limit is hardware-dependent. The source establishes specifications; it does not guarantee that maximum context fits at the LocalLens minimum RAM floor.