PID Intelligence Report — 2026-03-21 14:18 UTC

Status: YELLOW — 32/32 positions verified | 3 stale (>24h) | 11 changes in 24h


Arsenal Health

32/32 positions verified

ChainHealthNotes
I1 (Reasoning)11/11
I2 (General)11/13⚠ 2 unverified, 2 stale
I3 (Budget)10/10
Specialty4/4

⚠ Attention Required

PositionModelProviderIssueLast Verified
I2-pos5Mistral Large 3 675BNVIDIA (Key A)unverified, nevernever

Changes (Last 24h)

  • DeepSeek-R1-0528 — alert: PRICE DISCREPANCY: {‘openrouter’: 0.7, ‘artificial_analysis’: 1.35, ‘models_dev’: 0.574}
  • llama3.1-8b — alert: PRICE DISCREPANCY: {’litellm’: 0.03, ‘models_dev’: 0.1}
  • Meta-Llama-3.1-8B-Instruct — alert: PRICE DISCREPANCY: {’litellm’: 0.03, ‘models_dev’: 0.1}
  • deepseek-ai/deepseek-r1 — alert: PRICE DISCREPANCY: {‘openrouter’: 0.7, ‘artificial_analysis’: 1.35, ‘models_dev’: 0.574}
  • deepseek-r1-distill-llama-70b — alert: PRICE DISCREPANCY: {‘artificial_analysis’: 0.7, ‘models_dev’: 0.287}
  • deepseek-reasoner — alert: PRICE DISCREPANCY: {‘openrouter’: 0.7, ‘artificial_analysis’: 1.35, ‘models_dev’: 0.574}
  • qwen/qwen3.5-397b-a17b — alert: PRICE DISCREPANCY: {‘openrouter’: 0.39, ‘models_dev’: 0.6}
  • mistralai/mistral-large-3-675b-instruct- — alert: PRICE DISCREPANCY: {’litellm’: 0.5, ‘artificial_analysis’: 0.5, ‘models_dev’: 1.0}
  • llama-3.1-8b-instant — alert: PRICE DISCREPANCY: {’litellm’: 0.03, ‘models_dev’: 0.1}
  • openai/gpt-oss-120b:free — alert: PRICE DISCREPANCY: {’litellm’: 3.0, ‘openrouter’: 0.039}
  • openai/gpt-oss-120b — alert: PRICE DISCREPANCY: {’litellm’: 3.0, ‘openrouter’: 0.039}

Review Queue

Queue clear — no items pending.

10 items closed in last 48h.


Bench Depth

  • Active (Ring 1): 22 models — 10 reasoning — 5 tool-use — 16 large-context
  • Bench (Ring 2): 397 models — 4 reasoning — 3 tool-use — 7 large-context
  • Unsorted (Ring 3): 194 models — 84 reasoning — 146 large-context

Chain resilience: I1: 4 models across 7 providers, I2: 6 models across 8 providers, I3: 4 models across 9 providers


Opportunities

1 free reasoning models unassigned:

  • Nemotron 3 Super 120B (NVIDIA (Key A) — 262K ctx)

76 idle provider slots with verified capacity not in any chain.


Landscape

840 models tracked across 27 providers | Last scan: 15h ago (4587 models) Ring distribution: R1: 22 | R2: 616 | R3: 202

7-day activity: 11 changes across 1 active days


Generated 2026-03-21 14:18 UTC by pid-intelligence-report.py