[BOT ISSUE] fix: add missing Bedrock Qwen3 models (qwen3-235b, qwen3-vl-235b, qwen3-coder-next)

[github-actions]

Missing models

Qwen launched multiple models on Amazon Bedrock (announced September 2025 and later) but the catalog has zero qwen.* entries in packages/proxy/schema/model_list.json. This issue covers the three highest-impact Qwen models. Additional Qwen models (Qwen3 32B, Qwen3 Next 80B, Qwen3 Coder 30B/480B) also exist on Bedrock but are not included here to respect the 3-model limit per issue.

Model details

Field	qwen.qwen3-235b-a22b-2507-v1:0	qwen.qwen3-vl-235b-a22b	qwen.qwen3-coder-next
Format	converse	converse	converse
Flavor	chat	chat	chat
Multimodal	false	true (text + image)	false
Max input tokens	128000	256000	256000
Max output tokens	8000	8000	16000
Available providers	["bedrock"]	["bedrock"]	["bedrock"]
Display name	Qwen3 235B A22B	Qwen3 VL 235B A22B	Qwen3 Coder Next
Lifecycle	Active	Active	Active

Verification checklist

• Cross-source confirmation: All three models confirmed on (1) individual AWS Bedrock model cards (Qwen3 235B, Qwen3 VL 235B, Qwen3 Coder Next) with model IDs, context windows, and max output tokens, (2) AWS What's New announcement confirming GA availability, and (3) Qwen model family page on Bedrock listing all available models.
• Recent commits check: No qwen.* entries exist in model_list.json.
• ID format validation: Follow Bedrock naming convention (provider.model-name or provider.model-version:revision), consistent with existing entries.

Verification notes

Field	Source
Model IDs	Bedrock model cards — Programmatic Access tables
Context windows, max output	Bedrock model cards — Model Details sections
GA availability	AWS What's New
Qwen3 VL multimodal (text + image input)	Model card — Input Modalities table

Fields NOT verified from official sources:

• input_cost_per_mil_tokens / output_cost_per_mil_tokens — Pricing was not fully extractable from the Bedrock pricing page for all three models. Third-party sources suggest Qwen3 VL 235B is ~$0.11/$0.44 per MTok. The downstream fix job should retrieve the US standard tier pricing directly.
• reasoning — Qwen3 235B model description mentions "reasoning" capabilities but the model card doesn't explicitly document a reasoning token mode.
• Qwen3 VL and Qwen3 Coder Next also have bedrock-mantle Chat Completions API endpoints with potentially different model IDs (qwen.qwen3-vl-235b-a22b-instruct, qwen.qwen3-coder-next). The downstream fix job should consider these.

Additional Qwen models on Bedrock (not covered here)

These are also missing but excluded from this issue due to the 3-model limit:

• qwen.qwen3-32b — 32B dense model, 128K context
• qwen.qwen3-next-80b-a3b — 80B MoE with 3B active params
• qwen.qwen3-coder-30b-a3b-instruct — 30B MoE coding model
• qwen.qwen3-coder-480b-a35b-instruct — 480B MoE coding model

{
  "kind": "missing_model",
  "provider": "bedrock",
  "models": ["qwen.qwen3-235b-a22b-2507-v1:0", "qwen.qwen3-vl-235b-a22b", "qwen.qwen3-coder-next"],
  "status": "active",
  "model_specs": {
    "qwen.qwen3-235b-a22b-2507-v1:0": {
      "format": "converse",
      "flavor": "chat",
      "multimodal": false,
      "displayName": "Qwen3 235B A22B",
      "max_input_tokens": 128000,
      "max_output_tokens": 8000,
      "available_providers": ["bedrock"]
    },
    "qwen.qwen3-vl-235b-a22b": {
      "format": "converse",
      "flavor": "chat",
      "multimodal": true,
      "displayName": "Qwen3 VL 235B A22B",
      "max_input_tokens": 256000,
      "max_output_tokens": 8000,
      "available_providers": ["bedrock"]
    },
    "qwen.qwen3-coder-next": {
      "format": "converse",
      "flavor": "chat",
      "multimodal": false,
      "displayName": "Qwen3 Coder Next",
      "max_input_tokens": 256000,
      "max_output_tokens": 16000,
      "available_providers": ["bedrock"]
    }
  },
  "source_urls": [
    "https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-235b-a22b-2507.html",
    "https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-vl-235b-a22b.html",
    "https://docs.aws.amazon.com/bedrock/latest/userguide/model-card-qwen-qwen3-coder-next.html",
    "https://aws.amazon.com/about-aws/whats-new/2025/09/qwen3-models-fully-managed-amazon-bedrock/",
    "https://aws.amazon.com/bedrock/qwen/"
  ]
}