MMODELYST
Models/Llama 3.1 Instruct 70B
MDL

Llama 3.1 Instruct 70B

Meta

Verified against Artificial Analysis · Jul 13, 2026
Capability
38
percentile index
Price /M
$0.56 / $0.56
in / out
Context
Avg score
26.4
11 benchmarks
AA Intelligence
6.8
index
Speed
35 tok/s
TTFT 0.6s
Released
Jul 23, 2024
About

Llama 3.1 Instruct 70B is a open-weights model in the Llama family from Meta. Benchmarked on 11 evals, averaging 26.4.

Benchmark scores · 11
avg 26.4 — higher is better
MMLU-Pro
67.6
MATH
64.9
GPQA Diamond
40.9
IFBench
34.4
SciCode
26.7
LiveCodeBench
23.2
τ²-bench
15.2
AA Long-Context Reasoning
6.3
Humanity's Last Exam
4.6
AIME
4
Terminal-Bench Hard
3
Details
PublisherMeta
FamilyLlama
Parameters
License
WeightsOpen
ReleasedJul 23, 2024
Performance & price
Output speed35 tok/s
Latency (TTFT)0.6s
Price in / out$0.56 / $0.56 per 1M
Blended price$0.56 / 1M
Math index4.0

Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →

Capability over time

Append-only ledger — every observed change to this model's numbers.

API providers
Artificial Analysis$0.56/M