Llama 3.1-70B
Llama 3.1-70B is an open-weights AI model from Meta with an index score of 106 (GPT-4 = 100), rank 190 of 217.
10689–122Index score#190 of 217 models
$0.72/ $0.72Price per million tokensinput / output · via OpenRouter
131KContext windowtokens it reads at once
23 Jul 2024ReleasedReplaced by Llama 3.3 70B
Where it sitson the frontier
GPT-4 Turbo reached this score 9 months earlier.
Test resultsby area · tick = best by any model
Reasoning & knowledgeevidence
Mathsevidence
Codingevidence
Agentsevidence
Professional workno evidence yet
Novel problemsno evidence yet
Long tasksno evidence yet
Third-party results only, at the model’s best standard setting. Hover a bar for who ran it.
What others sayrank on 8 leaderboards
The Model Indexthird-party tests, our method#190of 217 · 106
LMArena · Textoverall chat quality, judged by people#240of 362 · 1293LMArena · Codingcoding questions in chat#238of 357 · 1333LMArena · Creative writingcreative writing#241of 360 · 1257Epoch Capabilities Indexoverall capability from many benchmarks#155of 231 · 124.7LMArena · Instruction followingfollowing detailed instructions#243of 362 · 1273LMArena · Mathsmaths questions in chat#233of 346 · 1269LMArena · Long promptslong, complex requests#239of 340 · 1295Best attop three among current models
Replaced by Llama 3.3 70B; it no longer leads any category among current models.
Where to use itprices checked 18 Nov 2026
API providers 1 · per 1M tokens, in / out
Run it yourself
Official weightsmeta-llama/Meta-Llama-3.1-70B-Instruct↗Host prices from OpenRouter; going direct can differ.
Facts and sources
Released23 Jul 2024
Announcementnot linked yet
Release entry–
Scored asLlama 3.1-70B
Output speed–
Hugging Facemeta-llama/Meta-Llama-3.1-70B-Instruct ↗