Meta: Llama 3.3 70B Instruct

Name: Meta: Llama 3.3 70B Instruct
Author: OpenRouter

70B

The Meta Llama 3.3 multilingual large language model (LLM) is a pretrained and instruction tuned generative model in 70B (text in/text out).

Context

131K

Max output

16K

Input price

$0.1/1M tokens

Output price

$0.32/1M tokens

Capabilities

vision

tool call

structured output

reasoning

json mode

streaming

fine tuning

batch

Details

Model IDmeta-llama/llama-3.3-70b-instruct

ProviderOpenRouter

Creatormeta-llama

Familyllama-3.3

License—

Parameters70B

Status—

Input modalitiestext

Output modalitiestext

Architecture—

Knowledge cutoff—

Training data cutoff—

Release date2024-12-06

Deprecation date—

Typechat

Reasoning tokens—

Max input—

Open weight—

Sourceofficial

Last updated2026-09-03

Tools

Function Callingfunction_calling

Call external functions and APIs

Pricing

Input

$0.1

↓86%$0.71

Output

$0.32

↓55%$0.71

Cache write

—

Cache read

$0.71

Batch in

—

Batch out

—

Price history · $/1M input

$0.1

$0.12

$0.1

$0.13

$0.1

$0.71

$0.1

Mar 21Apr 22Apr 22Jul 17Aug 5Aug 31Sep 3

Date	Input	Output	Cache read	Change
Mar 21, 2026	$0.1	$0.32	—	—
Apr 22, 20266231fb5	$0.12	$0.38	—	↑20%
Apr 22, 202680802be	$0.1	$0.32	—	↓17%
Jul 17, 202620c73ac	$0.13	$0.4	—	↑30%
Aug 5, 202619c9fe3	$0.1	$0.32	—	↓23%
Aug 31, 2026e3d7ded	$0.71	$0.71	$0.71	↑610%
Sep 3, 2026d79e481	$0.1	$0.32	$0.71	↓86%

Family Comparison: llama-3.3

Model	Caps	Perf	Speed	Context	Max out	Pricing
Meta: Llama 3.3 70B Instruct (free)	RVTS	—	—	131K	—
Meta: Llama 3.3 70B Instruct	RVTS	—	—	131K	16K	$0.1$0.32
NVIDIA: Llama 3.3 Nemotron Super 49B V1.5	RVTS	—	—	131K	16K	$0.4$0.4

API

GET/v1/models/openrouter/meta-llama/llama-3.3-70b-instruct