Build with Canadian AI
OpenAI-compatible inference API with full Canadian data residency. Content never touches US providers. Pay only for what you use.
from openai import OpenAI
client = OpenAI(
base_url="https://api.augureai.ca/v1",
api_key="your-api-key",
)
response = client.chat.completions.create(
model="ossington-5",
messages=[
{"role": "user", "content": "Summarize Bill C-27."}
],
)
print(response.choices[0].message.content)Three lines to switch
Already using the OpenAI SDK? Point your base URL at Augure, swap in your API key, and pick a model. Everything else — streaming, function calling, embeddings — works exactly the same.
Change your base URL
Point the OpenAI client at api.augureai.ca/v1. No new SDK required.
Use your Augure API key
Generate a key from the developer dashboard. Metered billing starts automatically.
Pick a model and ship
Choose Ossington 5 for fast reasoning and code, Rosedale One for the hardest work, or Tofino 3 for speed.
Sovereign infrastructure,
familiar interface
OpenAI Compatible
Drop-in replacement. Use the OpenAI SDK — just change the base URL and API key.
Canadian Sovereignty
All requests processed on Canadian-owned infrastructure. No US parent company in the stack.
Pay Per Use
Metered billing in CAD via Stripe. No minimum spend, no commitments.
Data Stored in Canada
Your data is stored in Canada; inference runs in Canada under zero retention — never US providers. Cross-border flows are documented in the privacy policy.
Simple, transparent rates
All prices in CAD per 1 M tokens. Ossington 5 cached input tokens: $0.60 per 1M. Other models bill cached input at the standard input rate.
Ossington 5
ossington-5Newest reasoning model. Fastest on the menu, strong at code.
Rosedale One
rosedale-1Premium reasoning for complex, multi-step work.
Tofino 3
tofino-3Light tier. Fast, thinking off, served in Canada with no failover outside Canada.
Augure Open
PreviewOpen models under their published names, hosted in Canada for a limited preview. Your messages reach the model as sent, text only, with no failover outside Canada. No SLA during the preview, and pricing may change.
GLM-5.3
open-glm-5-3Output up to 49,152 tokens. Thinking on by default.
Qwen3.8-Flash-Next
open-qwen-3.8-flash-nextOutput up to 32,768 tokens. Reasoning settings reach the model unchanged.
Ossington 4 ($2.00 / $5.00) and Ossington 4.1 ($2.50 / $6.00) remain available to existing integrations. See the model reference.
Augure OCR
augure-ocrScanned PDFs and images to Markdown, page by page, through POST /v1/ocr. Up to 30 pages per request. Served in Calgary, never outside Canada.
Powered by Augure
Applications using the Augure API must display “Powered by Augure” or the Augure logo in their customer-facing UI. We provide logos, badges, and embed snippets in the brand kit below.
Ready to build?
Sign up and start integrating in minutes.