Cerebras Pricing

Compare Cerebras inference pricing for 4 models. The world's fastest inference — up to 20x faster than alternatives.

Last updated: Feb 25, 2026

Models

~3,000

Fastest (tok/s)

$0.10

Cheapest Input/1M

Preview Models

Provider	Model	Speed (tok/s)	Context	Status	Input/1M	Output/1M
M Meta-llama	Llama 3.1 8B Instruct	~2,200	—	GA	$0.100	$0.100
O OpenAI	GPT-OSS-120b	~3,000	—	GA	$0.350	$0.750
QW Qwen	Qwen3 235B A22B Instruct 2507	~1,400	—	Preview	$0.600	$1.200
Z Z-ai	GLM 4.7	~1,000	—	Preview	$2.250	$2.750

Pricing from Cerebras. Prices per 1M tokens. Preview models (*) are not recommended for production use.