Pricing

Pricebook version 2026.10.8, effective 2026-10-01. All prices are in US dollars.

How you are billed

Never above list; your rate is shown before launch and frozen for the run. You pay what the provider bills for your machine, from the moment it is created until it is deleted, at provider cost ÷ 0.875 (Nodus keeps 12.5 % of what you pay). Each charge is itemized as Boot, Running, Restore or Teardown. Nodus-operated capacity bills at the node rates below. Credit is prepaid and every paid action reserves a hold first.

GPUs

List price per hour for the whole machine, by GPU count.

GPU	Memory	×1	×2	×4	×8
A10	24 GB	$0.89	$1.78	$3.56	$7.12
A100 40G	40 GB	$1.49	$2.98	$5.96	$11.92
A100 40G PCIE	40 GB	$1.39	$2.78	$5.56	$11.12
A100 80G	80 GB	$1.99	$3.98	$7.96	$15.92
A100 80G PCIE	80 GB	$1.89	$3.78	$7.56	$15.12
B200	180 GB	$6.49	$12.98	$25.96	$51.92
H100 PCIE	80 GB	$2.99	$5.98	$11.96	$23.92
H100 SXM	80 GB	$3.29	$6.58	$13.16	$26.32
H200	141 GB	$4.29	$8.58	$17.16	$34.32
L4	24 GB	$0.89	$1.78	$3.56	$7.12
L40S	48 GB	$1.29	$2.58	$5.16	$10.32
RTX 3090	24 GB	$0.59	$1.18	$2.36	$4.72
RTX 4090	24 GB	$0.69	$1.38	$2.76	$5.52
RTX 6000 ADA	48 GB	$1.09	$2.18	$4.36	$8.72
RTX A6000	48 GB	$0.79	$1.58	$3.16	$6.32

CPU machines

Shape	vCPU	Memory	Disk	Per hour
cpu-2x4	2	4 GiB	—	$0.16
cpu-2x8	2	8 GiB	—	$0.12
cpu-4x8	4	8 GiB	—	$0.31
cpu-4x16	4	16 GiB	—	$0.23
cpu-8x16	8	16 GiB	—	$0.62
cpu-8x32	8	32 GiB	—	$0.45
cpu-16x32	16	32 GiB	—	$0.79
cpu-16x64	16	64 GiB	—	$0.89
cpu-32x64	32	64 GiB	—	$1.58
cpu-32x128	32	128 GiB	—	$1.78
cpu-64x256	64	256 GiB	—	$3.56

Nodus nodes

Sandboxes, CPU Functions, CPU Workspaces and warm pools run on Nodus-operated nodes, billed per second from these rates.

Resource	Per hour
vCPU	$0.048
GiB of memory	$0.006
GiB of disk above 10 GiB per vCPU	$0.00014

Example	Per hour
Smallest Sandbox	$0.015
2 vCPU / 4 GiB Sandbox	$0.12
8 vCPU / 32 GiB CPU Job	$0.576

Storage and egress

Meter	Included	Price above the inclusion
Storage	10 GB per org	$0.017143 per GB-month
Egress	10 GiB per org per day	$0.09 per GiB
Relayed gang traffic	—	$0.011429 per GiB

Bring your own machines

Item	Price
Route: Nodus-scheduled work on your pool GPUs	$0.02 per device-hour
Predict	$99.00 per pool-month

Models

Hosted inference is billed per request from its usage.

Model	Operation	Price
baai/bge-m3	input	$0.013685 per 1M tokens
canopylabs/orpheus-v1-english	speech	$23.157895 per 1M characters
deepseek-ai/deepseek-v4-flash	input	$0.052632 per 1M tokens
deepseek-ai/deepseek-v4-flash	cacheRead	$0.046211 per 1M tokens
deepseek-ai/deepseek-v4-flash	output	$0.305264 per 1M tokens
deepseek-ai/deepseek-v4-pro	input	$0.410527 per 1M tokens
deepseek-ai/deepseek-v4-pro	cacheRead	$0.332527 per 1M tokens
deepseek-ai/deepseek-v4-pro	output	$3.673685 per 1M tokens
deepseek-ai/deepseek-v4.1-flash	input	$0.105264 per 1M tokens
deepseek-ai/deepseek-v4.1-flash	cacheRead	$0.047369 per 1M tokens
deepseek-ai/deepseek-v4.1-flash	output	$0.463158 per 1M tokens
moonshotai/kimi-k3	input	$1.51579 per 1M tokens
moonshotai/kimi-k3	cacheRead	$0.31579 per 1M tokens
moonshotai/kimi-k3	output	$9.473685 per 1M tokens
openai/gpt-oss-120b	input	$0.157895 per 1M tokens
openai/gpt-oss-120b	cacheRead	$0.078948 per 1M tokens
openai/gpt-oss-120b	output	$0.631579 per 1M tokens
openai/gpt-oss-20b	input	$0.052632 per 1M tokens
openai/gpt-oss-20b	cacheRead	$0.005264 per 1M tokens
openai/gpt-oss-20b	output	$0.210527 per 1M tokens
openai/whisper-large-v3	audio	$0.001948 per audio minute
openai/whisper-large-v3-turbo	audio	$0.000702 per audio minute
qwen/qwen3.8-27b	input	$0.442106 per 1M tokens
qwen/qwen3.8-27b	cacheRead	$0.089474 per 1M tokens
qwen/qwen3.8-27b	output	$3.157895 per 1M tokens
zai-org/glm-5.2	input	$0.242106 per 1M tokens
zai-org/glm-5.2	cacheRead	$0.196948 per 1M tokens
zai-org/glm-5.2	output	$4.631579 per 1M tokens
zai-org/glm-5.3	input	$0.231579 per 1M tokens
zai-org/glm-5.3	cacheRead	$0.186843 per 1M tokens
zai-org/glm-5.3	output	$3.568422 per 1M tokens
zai-org/glm-5.3-flash	input	$0.205264 per 1M tokens
zai-org/glm-5.3-flash	cacheRead	$0.041053 per 1M tokens
zai-org/glm-5.3-flash	output	$0.684211 per 1M tokens
nodus/indra	routing input	$0.044211 per 1M tokens
nodus/indra	routing output	$0.00 per 1M tokens

Plans

A plan’s allowance pays for its models at the prices above; it expires at the end of each month, and usage beyond it draws from your credits.

Plan	Per month	Allowance	Models
composer	$20.00	$20.00	nodus/indra

What you pay for

Item	Who pays
Your rented machine for every second the provider bills it, from creation to confirmed deletion: boot, image pull, restore, running, teardown and the provider’s rounding	You, at provider cost ÷ 0.875, never above list
Sandboxes, Functions, builds and CPU work on Nodus nodes, from placement to release	You, at the published node rates
Model tokens, including the routing call of nodus/auto	You, at the cheapest available model cost ÷ 0.95
Agent runs on Nodus-managed models	You, at model cost ÷ 0.875, from credits
Storage above the included amount and egress above the daily inclusion	You, at the published rates
Spare machines Nodus starts to finish sooner, and failures Nodus causes	Nodus
Requests whose outcome Nodus cannot confirm	Nodus
Logs, and egress within the daily inclusion	Nodus

Worked examples

Example	Cost
One H100 for one hour, at most	$3.29
A 2 vCPU / 4 GiB Sandbox for one hour	$0.12
1M output tokens on openai/gpt-oss-120b	$0.631579

Starter grant

The first org a verified user creates receives $30 of credit that expires 30 days after it is granted. Grant credit is spent before purchased credit.