modelbenchmark.io

Llama 3.1 70B Instruct

Meta · llama-3-1-70b

Compare

Open Llama instruction model for multilingual chat, reasoning, and coding

Specification

most-agreed values

Context
128K
Max output
8K
Released
2024-07-23
Knowledge cutoff
2023-12
Retires
Open weights
no
Input
text
Output
text

Price

US dollars per million tokens · most-agreed

Input
$0.72
Output
$0.72

Quality

3 benchmarks · 1 source

Composite
20th
percentile of 334 scored models
Rank
267 / 334
±0.06 sd
Evidence
3 × 1
one source, or fewer than 3 benchmarks
Effort range
one configuration only
reasoning25th1/4 bench
math17th2/7 bench

Every score

one row per source and configuration — nothing averaged away

BenchmarkScoreConfigurationSourceRun
MATH level 536.7 ±1.0defaultEpoch AI2025-01-27
GPQA diamond44.2 ±2.5defaultEpoch AI2025-01-27
OTIS Mock AIME 2024-20253.6 ±1.7defaultEpoch AI2025-02-25

The leading figure for a benchmark is the median across its configurations, so one heroic high-effort run cannot set the number.

Available from 2 hosts

HostIn $/MOut $/MCache rdCache wrContextOutputRetires
vercel$0.72$0.72128K8K
vercel_ai_gateway · vercel-ai-gateway$0.72$0.72128K8K

More from Meta

most-hosted first