[BOT ISSUE] fix: update stale pricing for xAI grok-4.20-0309-reasoning

[github-actions]

Stale pricing

The model grok-4.20-0309-reasoning has outdated pricing in the catalog (packages/proxy/schema/model_list.json, line 6801).

Current vs correct pricing

Field	Catalog value	Correct value
Input cost (per 1M tokens)	$2.00	$1.25
Output cost (per 1M tokens)	$6.00	$2.50

The catalog also has input_cache_read_cost_per_mil_tokens: 0.2, but the current xAI pricing page does not list a cached input price for this model. The downstream fix job should verify whether cached pricing still applies.

Context

The sibling model grok-4.3 is already correctly priced at $1.25/$2.50 in the catalog (line 7210), matching the current xAI pricing tier. The grok-4.20-0309-reasoning entry appears to have retained its original beta-era pricing while the official price has since been reduced to match the grok-4.3 tier.

Verification checklist

• Cross-source confirmation: Updated pricing confirmed on the xAI developer documentation page, which lists both the model and its pricing ($1.25/$2.50) in the same document. The pricing is consistent with the sibling grok-4.3 model already correctly priced in the catalog.
• Recent commits check: Pricing not corrected in any recent commits touching model_list.json.
• ID format validation: Model already exists in catalog (line 6801), ID format is correct.

Verification notes

Field	Source
Updated pricing ($1.25/$2.50)	xAI developer docs — pricing table
Model still active	xAI developer docs — model listing section

Local files inspected

• packages/proxy/schema/model_list.json — lines 6801–6814 show grok-4.20-0309-reasoning with input_cost_per_mil_tokens: 2 and output_cost_per_mil_tokens: 6

{
  "kind": "cost_update",
  "provider": "xAI",
  "models": ["grok-4.20-0309-reasoning"],
  "status": "active",
  "model_specs": {
    "grok-4.20-0309-reasoning": {
      "format": "openai",
      "flavor": "chat",
      "multimodal": true,
      "input_cost_per_mil_tokens": 1.25,
      "output_cost_per_mil_tokens": 2.5,
      "reasoning": true,
      "max_input_tokens": 2000000,
      "available_providers": ["xAI"]
    }
  },
  "source_urls": [
    "https://docs.x.ai/developers/models"
  ]
}