Best LLMs for Profiled Document Section Analysis

Category: Financial Analysis & Trading Decisions · Typical I/O: 8880→2085 tokens

Runs on model group: All models. Inputs may reach any model in the group, not only the ranked models below.

Run this task on Fronset — request an invitation

Models

Frontier on this task: Grok 4.7 at 9.85 / 10. Quality bar at 90%: 8.87.

point-estimate floor (CI low) · upper CI (less certain) · Bars sorted by blended cost; best-value model first. Greyed rows are MEDIUM+ models whose point estimate does not clear the bar.

Model	Quality score	CI low	Cost / 1k runs	vs best value
GPT-6 Luna	9.54 / 10	9.28	$0.83	best value
DeepSeek V4.1 Flash	9.03 / 10	8.71	$3.59	4.3x more expensive
GPT-6 Sol	9.02 / 10	8.77	$12.71	15x more expensive
GLM-5.3 Flash	9.09 / 10	8.91	$13.60	16x more expensive
Meta Muse Spark 1.3	8.88 / 10	8.64	$34.04	41x more expensive
Tencent Hy4 Preview	8.98 / 10	8.71	$40.59	49x more expensive
Grok 4.7	9.85 / 10	9.68	$54.04	65x more expensive
Moonshot Kimi K3	8.95 / 10	8.75	$71.53	87x more expensive
Thinking Machines Inkling Small	8.65 / 10	8.48	$8.74	11x more expensive
MiniMax M3	8.75 / 10	8.68	$3.30	4x more expensive
Tencent Hy3	7.15 / 10	6.66	$2.06	2.5x more expensive
Gemini 3.5 Flash Lite	7.09 / 10	6.79	$2.36	2.9x more expensive
Thinking Machines Inkling	8.77 / 10	8.63	$38.41	46x more expensive
Qwen 3.7 Plus	8.24 / 10	7.94	$7.17	8.7x more expensive
Claude Haiku 5.5	8.58 / 10	8.15	$4.84	5.9x more expensive
Claude Opus 5.5	8.33 / 10	7.90	$60.70	73x more expensive
GLM-5.3	8.76 / 10	8.49	$102.86	124x more expensive
Gemini 3.8 Flash	7.53 / 10	7.12	$6.92	8.4x more expensive
NVIDIA Nemotron-3 Super 120B	6.62 / 10	6.14	$2.40	2.9x more expensive
Xiaomi MiMo V2.6 Pro	8.86 / 10	8.42	$2.72	3.3x more expensive
NVIDIA Nemotron-3 Ultra 550B	7.85 / 10	7.46	$17.48	21x more expensive
Qwen 3.8 Max	8.34 / 10	7.92	$77.28	93x more expensive
Qwen 3.8 Flash	7.38 / 10	7.02	$15.77	19x more expensive
Claude Sonnet 5.5	8.48 / 10	8.07	$29.13	35x more expensive

Cost breakdown

Model	Quality	Confidence	Cost / 1k runs	Overpay	Mode
GPT-6 Luna ★ OpenAI	9.54 / 10 CI [9.28, 9.79]	HIGH	$0.83	best value	batch
DeepSeek V4.1 Flash DeepSeek	9.03 / 10 CI [8.71, 9.35]	MEDIUM	$3.59	4.3x	batch
GPT-6 Sol OpenAI	9.02 / 10 CI [8.77, 9.28]	HIGH	$12.71	15x	batch
GLM-5.3 Flash Z.AI	9.09 / 10 CI [8.91, 9.28]	RANKED	$13.60	16x	batch
Meta Muse Spark 1.3 OpenRouter	8.88 / 10 CI [8.64, 9.13]	HIGH	$34.04	41x	batch
Tencent Hy4 Preview OpenRouter	8.98 / 10 CI [8.71, 9.25]	HIGH	$40.59	49x	batch
Grok 4.7 best xAI	9.85 / 10 CI [9.68, 10.00]	RANKED	$54.04	65x	batch
Moonshot Kimi K3 Moonshot AI	8.95 / 10 CI [8.75, 9.14]	RANKED	$71.53	87x	batch

Overpay shows how much more you pay than the best-value model that clears the quality bar (marked ★) — the best-value good-enough option. "16x" means you overpay 16× — 16× that reference for no quality benefit above the bar. Typical call shape for this task: 8880 input tokens → 2085 output tokens, EMA-tracked from production traffic. Cost is the observed, all-in $ per 1,000 task runs: each model's own measured usage on this task — output verbosity, thinking/reasoning tokens, cache reads and writes, and the spend on its billed failures — priced at current list rates and adjusted by the billing overhead we actually reconcile against provider invoices. Models that answer tersely cost what they actually cost; models that think at length pay for it. Not comparable to providers' advertised $/1M list rates — this is what running the task costs, not a per-token price.

Evaluation rubric

Judge section fidelity, document-profile correctness, quantitative accuracy, materiality under the supplied lens, qualifier preservation, separation of fact and analysis, and recognition of missing context.

Output schema

Every answer on this task is checked against this JSON Schema, whichever model wrote it. An answer that doesn't fit counts as a model failure, and the call is retried on another model.

{
  "description": "Analysis result for a single chunk of a large SEC filing.",
  "properties": {
    "business_insights": {
      "title": "Business Insights",
      "type": "string"
    },
    "competitive_insights": {
      "title": "Competitive Insights",
      "type": "string"
    },
    "financial_insights": {
      "title": "Financial Insights",
      "type": "string"
    },
    "investment_relevance": {
      "title": "Investment Relevance",
      "type": "string"
    },
    "key_information": {
      "title": "Key Information",
      "type": "string"
    },
    "risk_insights": {
      "title": "Risk Insights",
      "type": "string"
    },
    "section_title": {
      "title": "Section Title",
      "type": "string"
    }
  },
  "required": [
    "section_title",
    "key_information",
    "business_insights",
    "financial_insights",
    "risk_insights",
    "competitive_insights",
    "investment_relevance"
  ],
  "title": "SecChunkAnalysis",
  "type": "object"
}

Prompt templates

The published template for this task.

default · LLMB_PROFILED_DOCUMENT_SECTION_ANALYSIS_SYSTEM + LLMB_PROFILED_DOCUMENT_SECTION_ANALYSIS_USER (601 calls in window)

System prompt

Analyze only the supplied section. Apply document_profile to interpret terminology, status, and qualifiers and analysis_profile to select material facts, metrics, risks, opportunities, and implications. Preserve whether figures are historical, adjusted, projected, or uncertain. State when a requested topic is outside the section rather than importing it from elsewhere. Treat empty optional values as absent and return only the requested result. Your response must conform exactly to this output schema: {schema_json_string}.

User prompt

Inputs — section_title: {section_title}; section_content: {section_content}; document_profile: {document_profile}; analysis_profile: {analysis_profile}. Use only these inputs to complete the task defined by the system prompt.