[BOT ISSUE] fix: add missing Fireworks models glm-5 and minimax-m2p5

[github-actions]

Missing models

Two Fireworks AI serverless models are listed on both the Fireworks models page and the Fireworks pricing page but are absent from the model catalog (packages/proxy/schema/model_list.json). The catalog already has closely related models from the same families (accounts/fireworks/models/glm-5p1, accounts/fireworks/models/minimax-m2p7).

Model details

accounts/fireworks/models/glm-5

Field	Value
Model ID	accounts/fireworks/models/glm-5
Format	openai
Flavor	chat
Input cost (per 1M tokens)	$1.00
Output cost (per 1M tokens)	$3.20
Cached input cost (per 1M tokens)	$0.20
Max input tokens	202752
Display name	GLM 5
Available providers	["fireworks"]

accounts/fireworks/models/minimax-m2p5

Field	Value
Model ID	accounts/fireworks/models/minimax-m2p5
Format	openai
Flavor	chat
Input cost (per 1M tokens)	$0.30
Output cost (per 1M tokens)	$1.20
Cached input cost (per 1M tokens)	$0.03
Max input tokens	196608
Display name	MiniMax M2.5
Available providers	["fireworks"]

Verification checklist

• Cross-source confirmation: Both models confirmed on (1) Fireworks models gallery page with pricing and context windows, and (2) Fireworks serverless pricing documentation page with input/output/cached pricing.
• Recent commits check: Neither model appears in any of the last 30 commits touching model_list.json.
• ID format validation: Follows existing accounts/fireworks/models/{name} convention used by sibling models accounts/fireworks/models/glm-5p1 (line 9746) and accounts/fireworks/models/minimax-m2p7 (line 10296).

Verification notes

Field	Source
Model IDs, context windows	Fireworks models page
Pricing (input/output/cached)	Fireworks serverless pricing docs
Format, flavor, provider convention	Inferred from sibling catalog entries glm-5p1 and minimax-m2p7

Fields NOT verified from official sources:

• max_output_tokens — not published on the models page or pricing page. The sibling glm-5p1 entry uses 128,000; minimax-m2p7 has no max_output_tokens set. The downstream fix job should verify via the API.
• multimodal — not confirmed; omitted.
• reasoning — the sibling glm-5p1 has reasoning: true, but this is not verified for glm-5. Omitted from spec.

{
  "kind": "missing_model",
  "provider": "fireworks",
  "models": ["accounts/fireworks/models/glm-5", "accounts/fireworks/models/minimax-m2p5"],
  "status": "active",
  "model_specs": {
    "accounts/fireworks/models/glm-5": {
      "format": "openai",
      "flavor": "chat",
      "input_cost_per_mil_tokens": 1.0,
      "output_cost_per_mil_tokens": 3.2,
      "input_cache_read_cost_per_mil_tokens": 0.2,
      "displayName": "GLM 5",
      "max_input_tokens": 202752,
      "available_providers": ["fireworks"]
    },
    "accounts/fireworks/models/minimax-m2p5": {
      "format": "openai",
      "flavor": "chat",
      "input_cost_per_mil_tokens": 0.3,
      "output_cost_per_mil_tokens": 1.2,
      "input_cache_read_cost_per_mil_tokens": 0.03,
      "displayName": "MiniMax M2.5",
      "max_input_tokens": 196608,
      "available_providers": ["fireworks"]
    }
  },
  "source_urls": [
    "https://fireworks.ai/models",
    "https://docs.fireworks.ai/serverless/pricing"
  ]
}