keiro 0.5.0

Keiro client — call the EB1 multi-model ensemble API.

Keiro

EB1 multi-model ensemble inference. Run multiple frontier models in parallel and synthesize the best response.

Quick start

pip install keiro
keiro setup

from keiro import models

print(models("eb1-preview", "What is machine learning?"))

How it works

EB1 sends your prompt to multiple frontier models (Claude, GPT, Gemini) in parallel, then a judge synthesizes the strongest elements into a single response. The result is more accurate and more complete than any individual model.

Models

Model	Description
eb1 (default)	3-model ensemble with synthesis
eb1-preview	Preview ensemble (GPT-5.2, Gemini, Claude)
eb1-pro	4-model ensemble for harder tasks
claude-opus-4-6	Direct passthrough (no ensemble)
gpt-5.2	Direct passthrough

from keiro import models

# Default ensemble
answer = models("eb1", "Solve this step by step: what is 23 * 47?")

# Specific model
answer = models("claude-opus-4-6", "Write a haiku")

Prompt-first API

from keiro import models

reply = models.response("eb1-preview", "Explain quantum computing in one paragraph.")
print(reply.text)

creative = models.instance("eb1-preview", temperature=0.8)
print(creative("Write a limerick about debugging."))

for chunk in models.stream("eb1-preview", "Draft a launch email."):
    print(chunk, end="")

complete(...) is still available as the smallest one-liner, but models is the preferred external interface because it matches Ember's public prompt-first API more closely.

Configuration

Interactive setup (recommended):

keiro setup

This validates your API key against the gateway and saves credentials to ~/.keiro/credentials.

Environment variables:

export KEIRO_API_KEY="your-api-key"
export KEIRO_BASE_URL="http://54.202.103.124:8080"  # optional

Explicit arguments:

from keiro import ModelsAPI

models = ModelsAPI(api_key="your-key", base_url="http://54.202.103.124:8080")
print(models("eb1-preview", "Hello"))

Precedence: explicit arguments > environment variables > credentials file.

Requirements

• Python 3.11+
• No GPU required (inference runs on Keiro's hosted infrastructure)