Feat: add io intelligence as a provider

[yug49]

Summary

Adds io.net Intelligence as a new provider

• Adds 7 model endpoints:
  • deepseek-ai/DeepSeek-R1-0528
  • deepseek-ai/DeepSeek-V3.2
  • mistralai/Mistral-Large-Instruct-2411
  • openai/gpt-oss-120b
  • zai-org/GLM-4.6
  • zai-org/GLM-4.7
  • meta-llama/Llama-4-Maverick-17B-128E-Instruct-FP8

[vercel]

@yug49 is attempting to deploy a commit to the Helicone Team on Vercel.

A member of the Team first needs to authorize it.

[yug49]

Hey team
I have a question before I can finalize this PR,

io.net provides an API enpoint, /api/v1/models that returns the complete list of available models with current pricing, context lengths, and other metadata. Since io.net frequently adds new models and updates pricing, static definitions in the codebase will quickly become outdated.

What's the recommended approach for keeping provider models auto-updated?

Happy to implement this if there's a preferred pattern or existing infrastructure.

[greptile-apps]

Your free trial has ended. If you'd like to continue receiving code reviews, you can add a payment method here.

[ilkhom19]

Hey team I have a question before I can finalize this PR,

io.net provides an API enpoint, /api/v1/models that returns the complete list of available models with current pricing, context lengths, and other metadata. Since io.net frequently adds new models and updates pricing, static definitions in the codebase will quickly become outdated.

What's the recommended approach for keeping provider models auto-updated?

Happy to implement this if there's a preferred pattern or existing infrastructure.

hi @yug49 , I am from IO Team, have you resolved this question ?

[yug49]

Hey team I have a question before I can finalize this PR,
io.net provides an API enpoint, /api/v1/models that returns the complete list of available models with current pricing, context lengths, and other metadata. Since io.net frequently adds new models and updates pricing, static definitions in the codebase will quickly become outdated.
What's the recommended approach for keeping provider models auto-updated?
Happy to implement this if there's a preferred pattern or existing infrastructure.

hi @yug49 , I am from IO Team, have you resolved this question ?

Hello @ilkhom19
Yes, the Helicone team was contacted and I was told that the best approach would be to go with the static values. This PR is ready to be reviewed and merged.

[ilkhom19]

LGTM