Llama 3.2 90B Instruct

Overview

Overview

Llama 3.2 90B is a large multimodal language model optimized for visual recognition, image reasoning, and captioning tasks. It supports a context length of 128,000 tokens and is designed for deployment on edge and mobile devices, offering state-of-the-art performance in image understanding and generative tasks.

Llama 3.2 90B Instruct was released on September 25, 2024. API access is available through 5 providers, including DeepInfra, Bedrock and others.

Benchmarks

Benchmarks

Llama 3.2 90B Instruct Performance Across Datasets

Scores sourced from the model's scorecard, paper, or official blog posts

Share

llm-stats.com - Sat Feb 21 2026

Pricing

Pricing

Pricing, performance, and capabilities for Llama 3.2 90B Instruct across different providers:

Provider	Input ($/M)	Output ($/M)	Max Input	Max Output	Latency (s)	Throughput	Quantization	Input	Output
DeepInfra	$0.35	$0.40	128.0K	128.0K	0.5	24.0 c/s	—	Text Image Audio Video	Text Image Audio Video
Bedrock	$0.72	$0.72	128.0K	128.0K	0.5	100.0 c/s	—	Text Image Audio Video	Text Image Audio Video
Fireworks	$0.89	$0.89	128.0K	128.0K	0.5	50.0 c/s	—	Text Image Audio Video	Text Image Audio Video
Together	$1.20	$1.20	128.0K	128.0K	0.5	57.0 c/s	—	Text Image Audio Video	Text Image Audio Video
Hyperbolic	$2.00	$2.00	128.0K	128.0K	0.5	42.0 c/s	—	Text Image Audio Video	Text Image Audio Video

Price Comparison for Llama 3.2 90B Instruct

Price per 1M input tokens (USD), lower is better

Share

llm-stats.com - Sat Feb 21 2026

Throughput Comparison for Llama 3.2 90B Instruct

Tokens per second, higher is better

Share

llm-stats.com - Sat Feb 21 2026

Latency Comparison for Llama 3.2 90B Instruct

Time to first token (s), lower is better

Share

llm-stats.com - Sat Feb 21 2026

API Access