Odysseus and billing API plan

[SirThinksAl0t]

Hi!
Any models I try to use, for example from OpenAI, it asks me for an active billing plan, error output here: (I tried different models too)
OpenAI rate-limited the request (429). You exceeded your current quota, please check your plan and billing details. For more information on this error, read the docs: https://platform.openai.com/docs/guides/error-codes/api-errors.

I thought this was all free?!? Or am I doing something wrong?!?
Does this mean anything I do it needs a billing API key?

[tronbonjovi]

Odysseus isn't a provider of AI, so it doesn't cost anything - but it also doesn't provide the AI service.

you can think of it like this:

the "brains" of the AI, is the model. that model runs somewhere, either on your computer (your gpu) or in a datacenter somewhere you access via the internet (cloud). there's apps you can use to run models on your own hardware (ollama, vllm, etc)

then there's the "body" or app that you use to do something with the model/brain (chat, generate an image write code, etc). this is like chatgpt/codex app/website or claude.

A lot of the cloud AI providers (openAI, anthropic, google, etc) provide service for their big models (brains), AND have their own app (body) to access their brain - chatgpt/codex, claude, gemini, etc.

Odysseus, is a "body" type app, it doesn't actually run the models itself, but it connects to the model and provides you the ability to interact with it. (it's also got some features to help with downloading the brains part, but that can be more complex)

[mikaelaldy]

Odysseus itself is completely free. But OpenAI's API is not free so you need to add a payment method and buy credits to use their models.

Your options:

Option 1: Pay for OpenAI

• Add a payment method at https://platform.openai.com/settings/organization/billing/overview
• Buy credits
• Gets you access to their models

Option 2: Use free models via OpenRouter
OpenRouter gives you access to free open-source models without any billing:

1. Get a free API key at https://openrouter.ai/settings/keys
2. In Odysseus, set endpoint to: https://openrouter.ai/api/v1
3. Use a free models

Heads up: Free models on OpenRouter are rate-limited to 50 requests/day (or 1,000/day if you later buy $10+ in credits). For experimenting, that's usually plenty

[SirThinksAl0t]

I checked out OpenRouter, but the first thing it says is "buy credits" and then it allows me to create a key... Where is the free section at?
Also tried & installed Ollama, but has no models. I need to create my own? or?
Sorry I'm new at this...

[mikaelaldy]

https://openrouter.ai/openrouter/free

[SirThinksAl0t]

I checked out Ollama, added a model, it took so much time saying in the console "processing first token/credit" (something like that), i thought it was gonna ask for money again. But no! It's working now!
Thank you guys for the explanation ❗
For the others wondering, I added:
ollama pull llama3.1

[tronbonjovi]

congrats! have fun!

a quick note -

you're running ollama on your machine - that means it's not sending your messages to a cloud server. it's private!
If you use any API (openrouter) - that's sending to a cloud server, it's not private.

just food for thought. you should NEVER send sensitive info to cloud models! :)

[skaraman]

scam detected