Anrak Ensemble: Fast
anrak-ensemble-fastanrak/ensemble-fastA single streamed pass on the platform's configured default model under your own system prompt. No fan-out, no consolidation, no verifier — this is the closest id in the catalog to calling an ordinary chat model, and it is priced accordingly. Reach for it when you want the AnrakLegal API surface and billing without any orchestration.
The Panel
This id runs a single pass — no panel convenes. One model answers, and nothing reviews it. For independent-error cancellation across four different model families, use anrak-ensemble.
Pricing
One flat charge per call, whatever the answer's length. A request to this id runs ~1 underlying model calls — that fan-out, not your token count, is what it costs us, so that is what we price on.
Estimate a month
Requests per month to anrak-ensemble-fast.
- Per request
- ₹12.32
- Per month
- ₹3,69,600
- Per 1,000 calls
- ₹12,320
Comparing against a per-token model?
On a reference request (2,000 tokens in, 2,000 out) this price works out to an effective $14.00 / $56.00 per 1M tokens.
That number looks high next to a raw frontier model, and it should: a raw model bills you for one call, while this id bills for 1 and a verification pass. We publish the comparison rather than hide it — but we bill per request, so a long answer does not cost you more.
Verification
No verifier
This id returns the model's answer directly. Nothing reviews it. That is the trade for the lowest latency and price in the catalog.
The verdict rides on every response as anrak.verify, the drafts and their grounded scores as anrak.orchestration, and the ordered trace as anrak.reasoning. When streaming, the same narration arrives live as delta.reasoning_content.
Quickstart
Two changes from an ordinary OpenAI call: the base_url and your ANRAK- key.
from openai import OpenAI
client = OpenAI(
base_url="https://anrak.legal/api/v1",
api_key="ANRAK-xxxxx",
)
resp = client.chat.completions.create(
model="anrak-ensemble-fast",
messages=[{"role": "user", "content": "Design a rate limiter for a multi-tenant API."}],
)
print(resp.choices[0].message.content)
# What it cost, and whether a downgrade happened:
extra = resp.model_extra or {}
print(extra["anrak"]["billing"]) # {'model': 'anrak-ensemble-fast', 'price': '₹12.32', ...}
print(extra["anrak"]["downgraded"]) # None unless the pipeline was downgradedLimits and trade-offs
- No ensemble and no verifier of any kind — the raw model's answer is returned.
- The underlying model is the platform's configured default, which we may re-point as models improve.
Best for
- • Cheap, fast general completions on the AnrakLegal surface
- • High-volume classification, extraction, and rewriting
- • A latency and price floor to compare the orchestrated ids against