Stop burning credits in Codex CLI. In this video I connect OpenAI Codex to OmniRoute — a free, open-source, MIT-licensed local AI gateway that pools 290+ providers and 90+ free tiers behind one endpoint — and get it running end to end in about 7 minutes.

OmniRoute runs on your own machine (localhost:20128), auto-falls back when a provider hits its quota, and stacks RTK + Caveman compression to cut token usage dramatically. Point Codex at one base URL and you get Claude, GPT, Gemini, Kimi, GLM, DeepSeek and hundreds more through a single config.

⚠️ Honest disclaimer: "unlimited" is shorthand for pooled free tiers (~1.4–1.5B tokens/month by OmniRoute's own pool-deduped count), not infinite tokens. Some providers restrict gateway use in their terms — check each provider's ToS before you route production work through it. This is an educational walkthrough of an open-source tool, not a way to bypass anyone's billing.

⏱️ CHAPTERS 00:00 — Why Codex eats your credits 00:45 — What OmniRoute actually is (290 providers, one endpoint) 01:30 — Install: npm i -g omniroute 02:20 — First run + the dashboard tour 03:10 — Connecting your first FREE providers 04:15 — Pointing Codex CLI at OmniRoute (config.toml) 05:20 — Testing it: model "auto" and auto-fallback in action 06:10 — Combos, routing strategies & token compression 06:50 — Costs, limits and what to watch out for

🔗 LINKS OmniRoute (GitHub): https://github.com/diegosouzapw/OmniRoute OmniRoute site: https://omniroute.online OpenAI Codex CLI: https://github.com/openai/codex

💬 Which provider combo are you running? Drop it in the comments — I'm collecting the best free-tier stacks for a follow-up video.

👍 Like + subscribe for more AI coding workflow videos.

#Codex #OmniRoute #AICoding #OpenSource #DeveloperTools

---

Transcript

0:00 Hello guys, welcome. I have this open AI

0:02 Codex open and it is working with Omni

0:05 route, which is a free API. They provide

0:07 you 1.53 billion tokens per month. Now,

0:11 if I ask this open AI Codex, which

0:14 models are you using? So, now you can

0:17 see it is using these models. From here,

0:21 you are using your models. Now, you can

0:24 see it is using these models and it is

0:27 using it from Integrity designed by

0:30 Google team. Now, if I come here to this

0:33 Omni route API, which I've already set

0:35 up, and if I come to these providers,

0:38 and now you can see here our Integrity

0:40 and Integrity CLI are connected. So, we

0:43 are using the models from this Omni

0:45 route API. So, how to set up this? In

0:48 this video, I'll show you step-by-step.

0:51 So, first we will open a new tab and

0:53 here I'll search for Omni route GitHub.

0:55 Now, go to this first link. Now, if I go

0:58 down, so here are the full details is

1:00 given. You can see here it has 290

1:02 providers, 90 plus is free, and it

1:05 provide 1.53 billion tokens per month

1:07 free. And for the first month, it

1:10 provide 2.15 billion tokens. How to set

1:13 up this? So, we will go down and here is

1:15 the quick start. Click on this. So, it

1:16 will take us here. So, to install Omni

1:19 route, you need NPM to be installed on

1:21 your system. So, for that, you need

1:23 Node.js. So, how can you install

1:24 Node.js? Search for Node.js and download

1:27 and the installation is pretty simple.

1:29 Next, next, next and it will install.

1:31 I'll copy this command from here. I'll

1:33 go here, open a new tab here, and I'll

1:36 paste it. Hit enter. And now it will

1:38 take some time to install this Omni

1:40 route. So, now Omni route is installed.

1:42 So, I'll go to this Omni route, which

1:43 I've already logged in, so I'll log out

1:45 from this. And now, I'll close this.

1:48 I'll come here again. I'll copy the

1:49 second command. And now, I'll paste it

1:52 here. Hit enter. And it will start Omni

1:54 route server and it will show you this

1:56 dashboard page. Now, the password is

1:58 right here. Copy this and paste it here.

2:01 Continue, and now you will log in to

2:04 Omni route. So, this is Omni route. So,

2:06 in this way, you can set up Omni route.

2:09 Now, we will install OpenAI Codex. So,

2:11 for that, I'll come to new tab, and

2:13 here, I'll search for Codex CLI. Go to

2:16 this first link. Go down. Select this

2:18 Codex CLI. Go down, and here, the

2:20 installation method is given. This is

2:22 for Mac Linux. So, click on this

2:24 Windows. Copy this. Come here. Open a

2:26 new window and paste the command here.

2:29 All right, enter. So, it will install

2:32 OpenAI Codex. So, now it will ask you to

2:35 start Codex. Put Y here and hit enter.

2:38 So, it will create all the files. Close

2:40 this window of Codex. Now, we will go

2:43 here to this folder. Open it. Go to this

2:45 PC, C, users. HP is my user. You will

2:49 have your own user. Now, you can see

2:51 here dot Codex folder. So, by default,

2:54 this is hidden. So, you can view it from

2:56 here. Show and hidden item. So, when you

2:59 check this, so it will show the hidden

3:01 folders. So, double click on this. Open

3:03 it. Here, search for config.toml.

3:06 It is right here. Open it. Maximize

3:08 this. So, you will not have all these

3:11 settings inside it because I've already

3:13 made the settings in it. So, I'll paste

3:15 these settings in the description of the

3:18 video. So, what you have to change are

3:20 the game start from here up to here. So,

3:23 if this is not available in your

3:25 config.toml

3:27 so you can copy it from here and paste

3:29 it there. Now, what is the base URL? So,

3:31 the base URL is from Omni route. So, for

3:33 that, I'll come here to Omni route. I'll

3:36 go to this endpoints, and this is our

3:38 base URL. So, I'll copy this, and we

3:41 will paste it right here. Now, this is

3:43 the API key. So, I'll generate a new API

3:45 key for you. For that, I'll come here

3:48 and go here and generate an API key.

3:51 I'll name this cordex, and create. I'll

3:54 copy the API key, and I'll paste it

3:56 here. Now, model provider is OpenAI, and

3:59 we will keep it auto. This will be auto.

4:02 I'll control S this. Now, we have to set

4:04 up some models. So, for that, I'll come

4:07 here. I'll click on this done, and now

4:09 I'll come to this providers. Now, here

4:11 we have the free tier. I've already set

4:13 up seven, but I'll show you how to set

4:15 up as this is showing error. So, I'll

4:17 click on this. I'll delete this, and

4:20 I'll add it again. Now, I'll type here

4:22 free in API key. Not everyone work with

4:25 this, but some of them work like this.

4:28 I'll save this, and now you can see two

4:30 models imported. I'll close this. I'll

4:33 go back. Now, I'll delete this entire

4:35 gravity as well, and we will connect it

4:38 again. Sign in. Copy this URL. Come

4:40 here, then to Omni route. Click on enter

4:43 manually, and paste the URL here. And

4:45 now, connect. So, it will connect entire

4:47 gravity CLI. I'll go again to providers.

4:50 Go to free tier. You can add all of

4:52 them, but some of them need API key. If

4:55 I click on this big box, add, and if I

4:58 type here free, and import free models,

5:01 and save, so now you can see two models

5:03 imported. I'll go back. So, all of these

5:06 are free tier models. Here, you can see

5:09 101 free tier models. You can add all of

5:12 them. Now, if I click on this deep seek,

5:14 and if I add, and if I use here free,

5:17 import, and save, so now you can see it

5:19 is saying invalid key. So, for this, you

5:22 have to go to their website and generate

5:25 the API key, and then add it here. So,

5:28 then it will start working. So, as I

5:30 have shown you how to add models, now we

5:33 are good to go. So, now I'll again go to

5:35 this CMD, and here I will open a new

5:38 tab, and now here I'll type for decks.

5:41 So, now you can see it is showing an

5:42 error because it is missing opening

5:45 quote. So, for that, I'll come here, and

5:47 we will put opening quote here. Control

5:49 S it. Now, I'll again we will go to CMD

5:51 and type codex and enter. So, now it is

5:54 working. So, now if I send it a hi, so

5:56 now you can see it has responded me,

5:58 "Hi, how can I help you?" Create me a

6:02 tic-tac-toe

6:03 game in HTML, CSS, and JavaScript and

6:08 save it in tic.html

6:12 all in one file. Hit enter. So, now you

6:15 can see it is working on it. So, we will

6:17 wait. So, now you can see it has started

6:19 working. I'll create a polished

6:21 tic-tac-toe game in a single HTML file

6:23 with embedded CSS and JavaScript. And it

6:26 has created the HTML file, but it's

6:28 still working on it. So, we will wait.

6:30 So, now you can see here your

6:32 tic-tac-toe game is ready. It has saved

6:34 it the file here. So, I'll copy this

6:36 path, and I'll go to this browser, open

6:38 a new tab, and paste the path here. So,

6:41 it will open the HTML file here as you

6:43 can see tic-tac-toe, and I play with X,

6:46 O, X, O, X. Now, you can see X won. One

6:50 If I use here, reset score. X, O, X, O,

6:55 X. Now, you can see again X won, and you

6:58 can create any round as well. So, in

7:01 this way you can set up codex CLI with

7:04 on the route, and you can use 1.53

7:07 billion tokens absolutely free. So, I

7:10 think that's all for today. I hope you

7:11 like the video. If you like the video,

7:13 please like, subscribe, and comment.

7:14 Thank you.