SDKs

Six SDKs, all OpenAI drop-in compatible.

Python - dash-py

59 tests. Idiomatic client with sync + async, typed dataclasses, and OpenAI drop-in examples including a RAG example showing the Claim + Evidence + Contradiction differentiator.

import openai

client = openai.OpenAI(
    base_url="http://localhost:8080/v1",
    api_key="not-needed",
)
response = client.embeddings.create(
    input="hello world",
    model="text-embedding-3-small",
)

Go - dash-go

86 tests. Go 1.21+ module with idiomatic functional options and the service pattern, full OpenAI drop-in compatibility, errors.Is/As support.

TypeScript - dash-ts

65 tests. ESM-first, Node 18+, zero runtime dependencies, full type safety with discriminated unions, drop-in OpenAI compatibility.

import OpenAI from 'openai';

const client = new OpenAI({
  baseURL: 'http://localhost:8080/v1',
  apiKey: 'not-needed',
});

const response = await client.embeddings.create({
  input: 'hello world',
  model: 'text-embedding-3-small',
});

Java, Kotlin, C#

Additional SDKs live in sdks/ - see the repository for status and examples.

Choosing

• Already on the OpenAI SDK? Point base_url at DASH - no new dependency needed.
• Need typed claim/evidence payloads? Use the native SDK for your language.
• Need contradiction filtering or temporal windows? Those live in the /v1/retrieve API, available from every SDK via the tenant-scoped client.