[BOT ISSUE] Bedrock: add missing pricing for 5 Qwen3 models

[github-actions]

Summary

Five Qwen3 Bedrock models were added to the catalog (via issues #549, #550, #629) but are missing input_cost_per_mil_tokens and output_cost_per_mil_tokens. AWS publishes on-demand pricing for all five on the Bedrock pricing page. Two sibling Qwen Bedrock entries (qwen.qwen3-235b-a22b-2507-v1:0 at $0.22/$0.88 and qwen.qwen3-coder-30b-a3b-v1:0 at $0.15/$0.60) already have correct pricing, confirming the pattern.

Affected Models

Model ID	Expected Input $/1M	Expected Output $/1M	Line
qwen.qwen3-32b-v1:0	0.15	0.60	~12458
qwen.qwen3-next-80b-a3b	0.15	1.20	~12470
qwen.qwen3-vl-235b-a22b	0.53	2.66	~12482
qwen.qwen3-coder-next	0.60	1.44	~12493
qwen.qwen3-coder-480b-a35b-v1:0	0.22	1.80	~12718

All five entries currently have available_providers: ["bedrock"], correct format, flavor, displayName, and token limits — only pricing fields are missing.

Verification

Check	Status	Detail
Cross-source (model existence)	✅	All 5 confirmed on Bedrock model cards AND Bedrock pricing page AND already present in model_list.json
Cross-source (pricing)	✅	US standard-tier pricing confirmed in closed issue #629 (cited Bedrock pricing page + third-party aggregators). Asia Pacific (Sydney) pricing independently confirmed via WebFetch ($0.1545/$0.6180 for Qwen3 32B — consistent with US $0.15/$0.60)
Sibling consistency	✅	Pricing aligns with already-priced siblings: qwen.qwen3-235b-a22b-2507-v1:0 ($0.22/$0.88), qwen.qwen3-coder-30b-a3b-v1:0 ($0.15/$0.60)
Standard tier available	✅	All model cards confirm Standard, Priority, and Flex tiers
Duplicate check	✅	No open issue covers Qwen Bedrock pricing. Prior issues (#549, #550, #629) that added the models are all closed

Verification Notes

Field	Source	Notes
Qwen3 32B pricing $0.15/$0.60	Issue #629 (US pricing), Bedrock pricing page (Sydney $0.1545/$0.6180)	Matches sibling qwen.qwen3-coder-30b-a3b-v1:0 pricing
Qwen3 Next 80B A3B pricing $0.15/$1.20	Issue #629 (US pricing)	Higher output price reflects larger model
Qwen3 VL 235B A22B pricing $0.53/$2.66	Issue #629 (US pricing), Bedrock pricing page (Sydney $0.2266/$0.9064 for non-VL variant)	Vision model premium over non-VL variant
Qwen3 Coder Next pricing $0.60/$1.44	Issue #629 (US pricing)	Latest coder model
Qwen3 Coder 480B A35B pricing $0.22/$1.80	Issue #629 (US pricing)	Matches input price of sibling 235B model

Note: The downstream fix job should verify exact US East (N. Virginia) pricing directly from the Bedrock pricing page, as the values above are from the closed issue #629 and Sydney-region WebFetch. The pricing page is dynamically rendered and may show different regions.

Local Files Inspected

• packages/proxy/schema/model_list.json — confirmed all 5 entries lack pricing fields
• packages/proxy/scripts/sync_models.ts — none of these models are in SYNC_PRESERVED_FIELDS

Proposed Changes

Add input_cost_per_mil_tokens and output_cost_per_mil_tokens to each entry.

{
  "kind": "cost_update",
  "provider": "bedrock",
  "models": ["qwen.qwen3-32b-v1:0", "qwen.qwen3-coder-480b-a35b-v1:0", "qwen.qwen3-coder-next"],
  "status": "needs_fix",
  "model_specs": {
    "qwen.qwen3-32b-v1:0": {
      "input_cost_per_mil_tokens": 0.15,
      "output_cost_per_mil_tokens": 0.60
    },
    "qwen.qwen3-coder-480b-a35b-v1:0": {
      "input_cost_per_mil_tokens": 0.22,
      "output_cost_per_mil_tokens": 1.80
    },
    "qwen.qwen3-coder-next": {
      "input_cost_per_mil_tokens": 0.60,
      "output_cost_per_mil_tokens": 1.44
    },
    "qwen.qwen3-next-80b-a3b": {
      "input_cost_per_mil_tokens": 0.15,
      "output_cost_per_mil_tokens": 1.20
    },
    "qwen.qwen3-vl-235b-a22b": {
      "input_cost_per_mil_tokens": 0.53,
      "output_cost_per_mil_tokens": 2.66
    }
  },
  "source_urls": [
    "https://aws.amazon.com/bedrock/pricing/",
    "https://docs.aws.amazon.com/bedrock/latest/userguide/model-cards-qwen.html"
  ]
}

[github-actions]

Autofix skipped because the issue body/title did not contain parseable bot metadata for a model catalog change.