Overview

---

Konko API is a fully managed API designed to help application developers:

1. Select the right LLM(s) for their application
2. Prototype with various open-source and proprietary LLMs
3. Access Fine Tuning for open-source LLMs to get industry-leading performance at a fraction of the cost
4. Setup low-cost production APIs according to security, privacy, throughput, latency SLAs without infrastructure set-up or administration using Konko AI's SOC 2 compliant, multi-cloud infrastructure

📘

The Konko API is fully compatible with the OpenAI API protocol

This allows the Konko API to be a drop-in replacement for applications currently using the OpenAI API but seeking open-source alternatives.

Guide: Switch OpenAI Apps to Konko API

How it works

---

To help you select the best LLM for your use-case: we provide free access to a variety of leading open and closed source models via our chat completion endpoint and completion endpoint. (more details on our supported LLMs here)

To request access to our free testing instance or custom fine-tuned models: email us at [email protected]. Please provide a brief description of your use case and we'll suggest the optimal model for your use case plus the requirements for additional fine-tuning.

Once you're ready to move to production, we provide a secure, low-cost API endpoint that meets your security, privacy, throughput, latency, and reliability requirements.

Moving to production is simple:

1. Submit your requirements through this form
2. Schedule a set-up call here

Once you've completed these two steps, the Konko AI team will setup access to your production-ready API.

Available Models

---

We recommend this curated list of industry-leading proprietary and open source LLMs for a variety of use cases. We offer fine-tuning services, hosting, or proxy access to the following models.

Chat Completion Endpoint

General Purpose

1. meta-llama/llama-2-70b-chat
2. meta-llama/llama-2-13b-chat
3. nousresearch/nous-hermes-llama2-13b
4. nousresearch/nous-hermes-llama-2-7b
5. mistralai/mistral-7b-instruct-v0.1
6. open-orca/mistral-7b-openorca

Expanded Context Window

1. konko/llama-2-7b-32k-instruct

Coding

1. codellama/codellama-34b-instruct
2. codellama/codellama-13b-instruct
3. codellama/codellama-7b-instruct

Completions Endpoint

General Purpose

1. meta-llama/llama-2-70b
2. meta-llama/llama-2-13b
3. mistralai/mistral-7b-v0.1

Expanded Context Window

1. konko/llama-2-7b-32k-instruct

Coding - General

1. codellama/codellama-34b
2. phind/phind-codellama-34b-v2

Coding - Python

1. codellama/codellama-34b-python
2. phind/phind-codellama-34b-python-v1

Coding - SQL

1. numbersstation/nsql-llama-2-7b

We're continuously evaluating and adding new models. If you'd like to see a new model, please request it by messaging [email protected] or joining our discord.

How to get started

---

First: get your API key. You can request it by messaging [email protected] or joining our discord.

Once you have your API key you can query the Konko API either:

1. Directly via HTTP in any popular language (learn how with our interactive documentation)
2. Via Konko AI's Python SDK (learn how with our Setup Access & SDK guide