Meta: Llama 4 Maverick
meta-llama/llama-4-maverick
Context: 1,048,576In: $0.20/1MOut: $0.80/1MUpdated 2026-08-05
Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward...
Intelligence
14.3
Coding
16.3
Agentic
1.3
Metric profile
Trend over time
Intelligence
14.3
Coding
16.3
Agentic
1.3
Only 1 snapshot so far (data refreshes daily) — the trend line appears after the 2nd snapshot.
Classic evals (curated from model cards)
Meta Llama — manually maintained, may lag releases
MMLU-Pro
72%
GPQA Diamond
60%
HumanEval
89%
Design Arena (ELO by category)
| Category | ELO | Win rate | Rank |
|---|---|---|---|
| 3d | 955 | 40.2% | #100 |
| uicomponent | 935 | 40.8% | #103 |
| codecategories | 909 | 35.8% | #112 |
| dataviz | 909 | 38.4% | #109 |
| website | 894 | 34.4% | #116 |
| gamedev | 883 | 33.7% | #110 |