[BOT ISSUE] Groq/Together: stale pricing for openai/gpt-oss-120b

[github-actions]

Gap

The openai/gpt-oss-120b entry in packages/proxy/schema/model_list.json has stale pricing. The catalog lists $0.10/$0.50 per million tokens (input/output), but both Groq and Together AI now charge $0.15/$0.60.

Catalog vs Official

Field	Catalog Value	Groq Official	Together Official
input_cost_per_mil_tokens	0.10	0.15	0.15
output_cost_per_mil_tokens	0.50	0.60	0.60

Affected Entry

• Model ID: openai/gpt-oss-120b
• File: packages/proxy/schema/model_list.json (line ~1532)
• Current providers: ["groq", "together", "baseten"]

Current catalog entry:

"openai/gpt-oss-120b": {
    "format": "openai",
    "flavor": "chat",
    "input_cost_per_mil_tokens": 0.1,
    "output_cost_per_mil_tokens": 0.5,
    "input_cache_read_cost_per_mil_tokens": 0.075,
    "displayName": "OpenAI GPT-OSS (120B)",
    "reasoning": true,
    "max_input_tokens": 131072,
    "max_output_tokens": 32766,
    "available_providers": ["groq", "together", "baseten"]
}

Verification

• Source 1: Groq models docs — production models pricing table lists openai/gpt-oss-120b at $0.15 input / $0.60 output per 1M tokens
• Source 2: Together AI serverless models — openai/gpt-oss-120b listed at $0.15 input / $0.60 output per 1M tokens

Both independent provider sources agree on the updated pricing, confirming the catalog value is stale.

Fields NOT changed

• input_cache_read_cost_per_mil_tokens (0.075) — not verified from current sources; may also need updating but not confirmed
• max_output_tokens (32766) — Groq lists 65,536 max output for this model. The catalog value of 32,766 may also be stale, but the Together docs did not publish max output, so this is noted but not filed as a confirmed gap
• Baseten pricing — not checked; the Baseten provider may have different pricing

Local files inspected

• packages/proxy/schema/model_list.json — confirmed current entry at line ~1532

Source URLs

• https://console.groq.com/docs/models
• https://docs.together.ai/docs/serverless-models

{
  "kind": "cost_update",
  "provider": "groq",
  "models": ["openai/gpt-oss-120b"],
  "status": "active",
  "model_specs": {
    "openai/gpt-oss-120b": {
      "format": "openai",
      "flavor": "chat",
      "input_cost_per_mil_tokens": 0.15,
      "output_cost_per_mil_tokens": 0.6,
      "displayName": "OpenAI GPT-OSS (120B)",
      "reasoning": true,
      "max_input_tokens": 131072,
      "max_output_tokens": 32766,
      "available_providers": ["groq", "together", "baseten"]
    }
  },
  "source_urls": [
    "https://console.groq.com/docs/models",
    "https://docs.together.ai/docs/serverless-models"
  ]
}

[github-actions]

Prepared catalog updates for openai/gpt-oss-120b. Updated models: openai/gpt-oss-120b.

Verification sources: https://console.groq.com/docs/models, https://docs.together.ai/docs/serverless-models
Verified metadata applied:

• openai/gpt-oss-120b: format=openai, flavor=chat, displayName=OpenAI GPT-OSS (120B), max_input_tokens=131072, max_output_tokens=65536, pricing(input/output per 1M)=0.15/0.6, input_cache_read_cost_per_mil_tokens=0.075, providers=groq|together|baseten, reasoning=true

Opened #576 to apply the catalog update.