Skip to content

Pricing

Simple, transparent pricing across all modalities.

Chat $0.06/1M tokens. Images $0.03 each. Audio $0.006/min. $10 free on signup — no credit card required.

Base rate

$0.06

CAD per 1M tokens. Input + output combined.

Free tier

$10

CAD free on signup. No credit card required.

Verification

Included

Signed receipts on every request at no extra cost.

Models

7

Chat, image, audio, and embedding models.

Chat completions

per 1M tokens

OpenAI-compatible chat inference with streaming support.

Price

$0.06 CAD / 1M tokens

Input + output tokens combined

Streaming and non-streaming

Signed receipt per request

Embeddings

per 1M tokens

Text embeddings for search, RAG, and classification.

Price

$0.06 CAD / 1M tokens

Single or batch input

all-MiniLM-L6-v2 default model

Signed receipt per request

Image generation

per image

Generate images with SDXL Turbo and Stable Diffusion XL on GPU nodes.

Price

$0.03 CAD / image

SDXL Turbo (fast) and SDXL (quality)

Up to 1024x1024 resolution

17x cheaper than DALL-E 3

Audio transcription

per minute

Transcribe audio with OpenAI Whisper. Same API format as OpenAI.

Price

$0.006 CAD / min

Whisper base and large-v3 models

Supports mp3, wav, m4a, flac, webm

Timestamps and language detection

Free tier

on signup

$10 in credits on every new account. No credit card required.

Price

$10 CAD free

Enough for ~170M tokens or 333 images

Start testing immediately

Additional credits via Stripe from $5

Volume

Need higher volume or custom models?

Get in touch for volume pricing, dedicated capacity, or custom model deployments.

What volume customers get

Pricing tied to committed monthly spend — not seat counts.

Option for dedicated GPU nodes with guaranteed availability.

Custom model deployments on the network.

Signed receipts and verification included at every tier.

Get started

Start with credits. When you need volume pricing or dedicated capacity, get in touch.

Operator economics

Transparent operator payouts

Operators earn from verified inference completions.

Earn from every verified inference completion on your GPU.
Payouts tied to signed receipts — proof of work, not uptime.
More models loaded = more jobs routed to your node.