Fronset Benchmark

Best LLMs for Authoritative Source Selection

Recommends authoritative, accessible, and maintainable sources for researching a subject, focus, regions, and optional analysis context. Illustrative uses include selecting sources for vendor diligence, market research, software standards, compliance monitoring, technology landsc

Cost mode:

Category: Relevance, Classification & Matching · Typical I/O: 10936→2921 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: GLM-5.3 at 8.65 / 10. Quality bar at 90%: 7.79.

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.

Cost breakdown

ModelQualityConfidenceCost / 1k runsOverpayMode
GLM-5.3 Flash ★ Z.AI8.11 / 10 CI [7.82, 8.40]HIGH$17.53best valuebatch
Meta Muse Spark 1.3 OpenRouter8.26 / 10 CI [8.07, 8.44]RANKED$34.952xbatch
Tencent Hy4 Preview OpenRouter8.13 / 10 CI [7.80, 8.46]MEDIUM$39.412.2xbatch
Claude Opus 5.5 Anthropic8.43 / 10 CI [8.24, 8.63]RANKED$63.613.6xbatch
Grok 4.7 xAI8.29 / 10 CI [7.99, 8.59]HIGH$67.353.8xbatch
Moonshot Kimi K3 Moonshot AI8.06 / 10 CI [7.58, 8.54]MEDIUM$86.114.9xbatch
GLM-5.3 best Z.AI8.65 / 10 CI [8.50, 8.81]RANKED$176.2510xbatch

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: 10936 input tokens → 2921 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 authority, directness, subject and regional relevance, expected accessibility, source-role diversity, URL/entity accuracy, and transparency about unverified assumptions.

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.

{
  "$defs": {
    "VettedSite": {
      "description": "Individual vetted site",
      "properties": {
        "category": {
          "default": "",
          "description": "Site category",
          "title": "Category",
          "type": "string"
        },
        "domain": {
          "description": "Domain name (e.g., example.com)",
          "title": "Domain",
          "type": "string"
        },
        "is_paywall_free": {
          "anyOf": [
            {
              "type": "boolean"
            },
            {
              "type": "null"
            }
          ],
          "default": null,
          "description": "Whether the site is paywall-free",
          "title": "Is Paywall Free"
        },
        "name": {
          "description": "Site name",
          "title": "Name",
          "type": "string"
        },
        "reasoning": {
          "default": "",
          "description": "Why this site is relevant",
          "title": "Reasoning",
          "type": "string"
        }
      },
      "required": [
        "domain",
        "name"
      ],
      "title": "VettedSite",
      "type": "object"
    }
  },
  "description": "Output model for vetted site selection",
  "properties": {
    "vetted_sites": {
      "description": "List of selected vetted sites",
      "items": {
        "$ref": "#/$defs/VettedSite"
      },
      "title": "Vetted Sites",
      "type": "array"
    }
  },
  "required": [
    "vetted_sites"
  ],
  "title": "VettedSitesOutput",
  "type": "object"
}

Prompt templates

The published template for this task.

default · LLMB_AUTHORITATIVE_SOURCE_SELECTION_SYSTEM + LLMB_AUTHORITATIVE_SOURCE_SELECTION_USER (0 calls in window)

System prompt

Select sources likely to provide direct, current, and citable evidence for the stated research focus. Prioritize official, primary, regulatory, academic, or established specialist sources as appropriate. Explain each source’s evidence role and regional relevance. Do not invent URLs, accessibility, update frequency, or authority; exclude aggregators and low-signal sources when stronger primary sources are available. 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 — subject_name: {subject_name}; subject_type: {subject_type}; subject_description: {subject_description}; research_focus: {research_focus}; chapter_context: {chapter_context}; regions: {regions}; additional_requirements: {additional_requirements}. Use only these inputs to complete the task defined by the system prompt.