← Leaderboard

DeepSeek-R1-Distill-Llama 70B 8bit MLX

Retired Rank #30 of 33 · 2/8 GAUNTLET progress
Generalist — 58 / 100 G Agentic — pending A Understanding — pending U Needle — pending N Thinking — pending T Live — pending L Engineering — pending E Throughput — 3.5 / 100 T

Hollow markers & dashed spokes: axis not yet scored

G
58
A
U
N
T
n/d
L
E
T
3.5

Specification

Parameters
70B
Architecture
llama
Size on disk
75 GB
Quantization
8bit
Format
MLX
Reasoning (CoT)
Yes — emits reasoning tokens
Internal ID
M5
Mean speed
5.7 tok/s across suites
Stall census
0 stalls in 13 observed tests (0.0%)
Reasoning appetite
1,049 tokens mean · 6,741 max

Suite results

Suite Score Avg / 20 Tests tok/s
General capability (13-task real-workload suite) 1a 150.8 / 260 11.6 13/13 5.7

Each test is LLM-judged 0–20 against a fixed rubric; suite max = tests × 20. Where fewer tests ran than the suite total, unrun tests count as zero toward the axis score. See methodology.