Frontier open models, fine-tunes, custom weights. If it can run, we serve it.
Edge routing, regional clusters, and tiered capacity keep latency low at any scale.
Same endpoints. Same request and response shape. Switch by changing the base URL.
Server-sent events, tool use, function calling, structured outputs. All first-class.
Every call produces a signed receipt. Output hashes recorded on-chain.
Compute where your users are. Sub-50ms anywhere with a network connection.
Token-efficient, with volume discounts and burn incentives.
Input tokens. Output at 3× input.
Chain-of-thought, longer context.
Up to 1024×1024 resolution.
Batched at 256 texts/request.
47 voices, 31 languages.
Speaker diarization included.
Pricing is illustrative. Final rates are governed by on-chain parameters and may vary based on network state.
Tier-based access. Higher stakes unlock better economics and more capacity.
| Tier | Required stake | Access |
|---|---|---|
| Explorer | 10M QUBIC | Best-effort, shared pool |
| Builder | 50M QUBIC | Reserved throughput, 25% burn discount |
| Startup | 150M QUBIC | Dedicated cluster, 40% burn discount |
| Business | 500M QUBIC | Enterprise SLA, 55% burn discount |
| Enterprise | Custom | Custom infrastructure and SLAs |
Drop-in compatible with the OpenAI SDK.
from aigarth import Aigarth
client = Aigarth(api_key="sk-aigarth-...")
response = client.chat.create(
model="aigarth-reason-1",
messages=[
{"role": "system", "content": "You are a helpful assistant."},
{"role": "user", "content": "What's the capital of France?"},
],
stream=True,
)
for chunk in response:
print(chunk.choices[0].delta.content or "", end="")Everything in the standard tier, plus the things enterprises need.