Kilo Code

Configure Kilo Code, the open-source AI coding agent for VS Code and the terminal, to use Eden AI for 500+ models behind one API key.

Overview

Eden AI is a built-in provider in Kilo Code, in both the VS Code extension and the CLI. Kilo uses the provider ID edenai and reads your key from EDENAI_API_KEY.

500+ models: reach OpenAI, Anthropic, Google, Mistral, Z.ai and more through one key
One key, one invoice: no per-vendor accounts to manage
EU endpoint: route through api.eu.edenai.run when you have data-residency requirements
Shared credential store: the extension and the CLI read the same auth.json

Prerequisites

Kilo Code installed, either the VS Code extension or the CLI. See kilocode.ai
An Eden AI API key from app.edenai.run → API Keys

Setup in VS Code

Open Settings in the Kilo Code extension.
Go to the Providers tab and add Eden AI. If it is not listed, click Show more providers.
Paste your Eden AI API key.
Select a model.

Credentials are stored in Kilo’s auth.json, the same store the CLI uses.

Setup in the CLI

Connecting interactively is the recommended path, so the key lands in auth.json rather than in a project file. In the TUI run /connect, choose Eden AI, then paste your key. From the shell:

kilo auth login --provider edenai

Then pick a model from the picker, or set a default:

{
  "model": "edenai/anthropic/claude-sonnet-5"
}

To keep the secret in the environment instead, declare the provider in ~/.config/kilo/kilo.json or ./kilo.json:

export EDENAI_API_KEY="your_api_key_here"

{
  "provider": {
    "edenai": {
      "env": ["EDENAI_API_KEY"]
    }
  },
  "model": "edenai/anthropic/claude-sonnet-5"
}

Model naming

Eden AI model IDs are already vendor/model, so a full reference in Kilo has three segments, edenai/<vendor>/<model>:

{ "model": "edenai/openai/gpt-4o-mini" }
{ "model": "edenai/google/gemini-3.6-flash" }
{ "model": "edenai/zai/glm-5.2" }

Copy the identifier exactly as the models endpoint returns it. It is public and needs no key:

curl https://api.edenai.run/v3/models

Using a model that is not in the picker

The picker lists the Eden AI models in the catalogue Kilo refreshes, which is smaller than the full Eden AI catalogue. Any other model can be declared yourself:

{
  "provider": {
    "edenai": {
      "models": {
        "anthropic/claude-haiku-4-5": {
          "name": "Claude Haiku 4.5",
          "cost": { "input": 1, "output": 5, "cache_read": 0.1 }
        }
      }
    }
  },
  "model": "edenai/anthropic/claude-haiku-4-5"
}

Set cost when you declare a model, and limit too if you want the context gauge to be right. Kilo defaults both to zero for a model it does not already know, which would leave spend reporting at zero. Costs are per million tokens, so multiply the endpoint’s input_cost_per_token and output_cost_per_token by one million. A limit block needs both context, from context_length, and output.

EU data residency

Point the provider at the EU gateway with a custom base URL:

{
  "provider": {
    "edenai": {
      "env": ["EDENAI_API_KEY"],
      "options": {
        "baseURL": "https://api.eu.edenai.run/v3"
      }
    }
  }
}

That base URL governs where Eden AI processes the request. The EU endpoint serves the subset of the catalogue available in the EU, so a model that works on the global endpoint is not guaranteed to be reachable through it. Independently, some models are published as a region-pinned variant that also pins the underlying vendor’s region, marked with an @eu suffix:

{ "model": "edenai/vertex/gemini-3.6-flash@eu" }

Only a few of these appear in the picker. Declare the others as shown in the previous section.

Cost tracking

The spend Kilo Code displays for Eden AI is derived from token counts and the per-model rates in its catalogue, so treat it as an estimate. Eden AI returns the actual billed cost of each request in its own response body, and that figure, with your Eden AI dashboard, is the authoritative one.

Troubleshooting

Invalid API key

Create a new key in your Eden AI account, then reconnect with /connect or kilo auth login --provider edenai. If you use manual configuration, update EDENAI_API_KEY in the environment Kilo runs from.

Model not found

Copy the identifier exactly as the models endpoint returns it, vendor prefix included. Do not rely on a static list.

Eden AI is not visible in the provider list

Refresh Kilo Code’s provider catalogue, then check Show more providers.

A key or model you just added has no effect

Kilo caches the resolved catalogue for a few minutes. Wait, then restart your client.

Next Steps

Chat Completions - Core LLM endpoint
List LLM Models - Browse available providers and models
Cline - Another VS Code coding agent on Eden AI
OpenCode - Terminal coding agent on Eden AI

V3 Documentation

Quick Start

Overview

LLMs

Expert Models

General

Organization

Data Governance

Integrations

Overview

Prerequisites

Setup in VS Code

Setup in the CLI

Model naming

Using a model that is not in the picker

EU data residency

Cost tracking

Troubleshooting

Invalid API key

Model not found

Eden AI is not visible in the provider list

A key or model you just added has no effect

Next Steps

​Overview

​Prerequisites

​Setup in VS Code

​Setup in the CLI

​Model naming

​Using a model that is not in the picker

​EU data residency

​Cost tracking

​Troubleshooting

​Invalid API key

​Model not found

​Eden AI is not visible in the provider list

​A key or model you just added has no effect

​Next Steps

Overview

Prerequisites

Setup in VS Code

Setup in the CLI

Model naming

Using a model that is not in the picker

EU data residency

Cost tracking

Troubleshooting

Invalid API key

Model not found

Eden AI is not visible in the provider list

A key or model you just added has no effect

Next Steps