chore: configs for lightning-ai provider

[whysumedh]

No description provided.

[Copilot]

Pull request overview

This PR adds a new provider entry for lightning-ai by introducing pricing data and general capability/parameter configuration so the models can be surfaced consistently alongside existing providers.

Changes:

• Added pricing/lightning-ai.json with default pricing calculation/currency and per-model pay-as-you-go rates.
• Added general/lightning-ai.json with provider defaults (params/message roles) and per-model capability flags (e.g., image support).

Reviewed changes

Copilot reviewed 2 out of 2 changed files in this pull request and generated 1 comment.

File	Description
pricing/lightning-ai.json	Introduces Lightning AI pricing configuration (default calculation + model rate overrides).
general/lightning-ai.json	Introduces Lightning AI model capability/parameter configuration (defaults + per-model type/supported flags).

Comments suppressed due to low confidence (4)

general/lightning-ai.json:120

• For openai/o3-mini, the entry currently has no model-specific max_tokens limit and does not advertise tool support. Other provider-aggregator configs in this repo set max_tokens up to 100000 and mark tool support for this model; without this, clients may be constrained incorrectly or miss supported features.

  "openai/o3-mini": {
    "type": { "primary": "chat", "supported": [] }
  },

general/lightning-ai.json:123

• For openai/gpt-4o, the config currently only lists image support and inherits the default max_tokens limit. In other configs in this repo this model supports tools and has a smaller max_tokens cap; consider explicitly setting these to avoid clients sending out-of-range values.

  "openai/gpt-4o": {
    "type": { "primary": "chat", "supported": ["image"] }
  },

general/lightning-ai.json:129

• For openai/gpt-4-turbo-preview, the config currently has an empty supported list and inherits the default max_tokens limit. This model typically supports tool calling and a lower max_tokens cap; without overrides, clients may attempt unsupported parameter values/features.

  "openai/gpt-4-turbo-preview": {
    "type": { "primary": "chat", "supported": [] }
  },

general/lightning-ai.json:132

• For openai/gpt-3.5-turbo, the entry currently omits tool support and inherits the default max_tokens maxValue (32768). Other configs in this repo constrain this model to a 4096 max_tokens limit and mark tool support; consider aligning to prevent clients sending unsupported requests.

  "openai/gpt-3.5-turbo": {
    "type": { "primary": "chat", "supported": [] }
  },

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.