Meta AI · model passport

Llama 4 Maverick 17b 128e Instruct Fp8

llama-4-maverick-17b-128e-instruct-fp8

Best percentile: 62.9HHEM Hallucination (Vectara)
Composite standing: 44.8shrunk · 5/27 axes

What the data says

Benchmarked on 5 of 27 tracked axes. Every figure below is the model's own published result on that board.

Weakest axes

11.6 pct — FrontierMath (0.7%)

Uneven: 51.3 percentile points between its best and worst axis — 62.9 pct on HHEM Hallucination (Vectara) versus 11.6 pct on FrontierMath.

HHEM Hallucination (Vectara)62.9 pct91.8%Verify →
BFCL v4 (function calling)Dormant board56.1 pct37.3%Verify →
GPQA Diamond48.9 pct67.0%Verify →
AIME (math)29 pct20.6%Verify →
FrontierMath11.6 pct0.7%Verify →
OverallNot benchmarked
WebDev ArenaNot benchmarked
Vision ArenaNot benchmarked
Factuality ArenaNot benchmarked
Search ArenaNot benchmarked
Document ArenaNot benchmarked
Agent ArenaNot benchmarked
SimpleQA VerifiedNot benchmarked
ARC-AGI-2Not benchmarked
Humanity's Last ExamNot benchmarked
SimpleBenchNot benchmarked
LiveBenchNot benchmarked
SWE-bench Verified (live board)Not benchmarked
Terminal-Bench 4.0Not benchmarked
MMLU-Pro (Artificial Analysis)Not benchmarked
LiveCodeBench (Artificial Analysis)Not benchmarked
IFBench (Artificial Analysis)Not benchmarked
SWE-bench VerifiedNot benchmarked
Terminal-BenchNot benchmarked
CodingNot benchmarked
Instruction followingNot benchmarked
Hard promptsNot benchmarked

Percentiles are the model's standing within each board's own population. Axes marked "not benchmarked" have no published result for this model — that is an absence of data, not a low score.

Llama-4-Maverick-17B-128E-Instruct-FP8 stands at 44.8 (raw mean 41.7) — measured on 5 of 27 axes.
standingtemplateVerify →

Every claim is a computed fact from the current board data, phrased by a low-temperature model call and rejected by a validation gate if any number is not in the source row. How the leaderboards work →