All models
Meta

Llama 3.3 70B

standardOpen Source

Meta

Meta's high-performance open-source model. Near frontier quality in a 70B package — fully free to download, self-host, and fine-tune.

Official page
open-sourceself-hostablefine-tuning
Context
131.072K
Input / 1M
$0.10
Output / 1M
$0.32
Knowledge cutoff
2023-12

Capabilities

Code4/5
Reasoning3/5
Multilingual3/5
Safety3/5
Speed4/5

Modalities & Features

Vision (image input) No
Audio No
Video No
Function / tool calling Yes
JSON mode Yes
Extended thinking No

Context & Output

Context window131.072K
Max output tokens8.192K
Knowledge cutoff2023-12
Release date2024-12

API & Access

Open source Yes
Streaming Yes
Batch API No
Prompt caching No
Fine-tuning Yes

Benchmarks

MMLUGeneral knowledge
86%
GPQA DiamondHard science Q&A
50.5%
MATHMath problem solving
73.5%
AIME 2025Advanced math
N/A
GSM8KGrade-school math
93.7%
HumanEvalCode generation
78.9%
LiveCodeBenchLive coding problems
N/A
SWE-bench VerifiedReal-world software
N/A
HellaSwagCommonsense inference
N/A

Source: provider technical reports & independent evaluations. N/A = not yet published for this model.

Pricing (per 1M tokens)

Input$0.10
Output$0.32
Prices shown are list prices and may vary by provider plan. Always check official documentation for the latest pricing.