Best LLMs for Generated Content Relevance Scoring

Category: Relevance, Classification & Matching · Typical I/O: 12037→250 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: Qwen 3.7 Plus at 8.95 / 10. Quality bar at 90%: 8.05.

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
DeepSeek V4.1 Flash	8.59 / 10	8.24	$1.58	best value
Xiaomi MiMo V2.6 Pro	8.43 / 10	8.01	$2.19	1.4x more expensive
GPT-6 Luna	8.42 / 10	7.96	$3.77	2.4x more expensive
Gemini 3.5 Flash Lite	8.18 / 10	7.71	$6.07	3.8x more expensive
Qwen 3.7 Plus	8.95 / 10	8.67	$9.56	6.1x more expensive
GPT-6 Sol	8.35 / 10	7.87	$14.13	8.9x more expensive
Thinking Machines Inkling Small	8.23 / 10	7.77	$14.15	9x more expensive
Grok 4.7	8.77 / 10	8.49	$32.42	21x more expensive
Claude Opus 5.5	8.47 / 10	8.07	$58.24	37x more expensive
Tencent Hy3	8.02 / 10	7.54	$4.46	2.8x more expensive
NVIDIA Nemotron-3 Ultra 550B	7.88 / 10	7.47	$11.38	7.2x more expensive
NVIDIA Nemotron-3 Super 120B	7.91 / 10	7.50	$3.76	2.4x more expensive

Cost breakdown

Model	Quality	Confidence	Cost / 1k runs	Overpay	Mode
DeepSeek V4.1 Flash ★ DeepSeek	8.59 / 10 CI [8.24, 8.93]	MEDIUM	$1.58	best value	batch
Xiaomi MiMo V2.6 Pro Xiaomi MiMo	8.43 / 10 CI [8.01, 8.86]	MEDIUM	$2.19	1.4x	batch
GPT-6 Luna OpenAI	8.42 / 10 CI [7.96, 8.88]	MEDIUM	$3.77	2.4x	batch
Gemini 3.5 Flash Lite Gemini	8.18 / 10 CI [7.71, 8.65]	MEDIUM	$6.07	3.8x	batch
Qwen 3.7 Plus best Alibaba Cloud (DashScope)	8.95 / 10 CI [8.67, 9.23]	HIGH	$9.56	6.1x	batch
GPT-6 Sol OpenAI	8.35 / 10 CI [7.87, 8.83]	MEDIUM	$14.13	8.9x	batch
Thinking Machines Inkling Small OpenRouter	8.23 / 10 CI [7.77, 8.69]	MEDIUM	$14.15	9x	batch
Grok 4.7 xAI	8.77 / 10 CI [8.49, 9.05]	HIGH	$32.42	21x	batch
Claude Opus 5.5 Anthropic	8.47 / 10 CI [8.07, 8.88]	MEDIUM	$58.24	37x	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: 12037 input tokens → 250 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 one-to-one coverage of the supplied set, semantic relevance to the supplied context, recognition of material contribution or necessary qualification, score and threshold calibration, consistency across items, preservation of item identity, separation of content quality from contextual fit, and resistance to label, length, polish, and neighboring-item bias.

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.

{
  "$defs": {
    "RelevanceScoreItem": {
      "description": "Generic relevance score item for any source/target pair.",
      "properties": {
        "id": {
          "description": "ID of the item being scored",
          "title": "Id",
          "type": "string"
        },
        "relevance_score": {
          "description": "Relevance score 0.0-1.0",
          "maximum": 1,
          "minimum": 0,
          "title": "Relevance Score",
          "type": "number"
        }
      },
      "required": [
        "id",
        "relevance_score"
      ],
      "title": "RelevanceScoreItem",
      "type": "object"
    }
  },
  "description": "Generic LLM output for relevance scoring.",
  "properties": {
    "items": {
      "items": {
        "$ref": "#/$defs/RelevanceScoreItem"
      },
      "title": "Items",
      "type": "array"
    }
  },
  "required": [
    "items"
  ],
  "title": "RelevanceScoreResult",
  "type": "object"
}

Prompt templates

The published template for this task.

default · LLMB_GENERATED_CONTENT_RELEVANCE_SCORING_SYSTEM + LLMB_GENERATED_CONTENT_RELEVANCE_SCORING_USER (2035 calls in window)

System prompt

Evaluate every supplied item for substantive relevance to the supplied scoring context. The context may describe an overall objective, question, scope, audience need, or destination section specification. Determine whether each item materially advances that context through evidence, analysis, useful background, or necessary qualification. Score items independently based on their actual content, not their label, length, polish, or similarity to neighboring items. A high-quality item that does not help satisfy the context must score low. Preserve each item identifier exactly, return one result for every input item, and apply the same relevance scale consistently across the set. 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 — scoring_context: {scoring_context}; items_list: {items_list}.

Relevance scale — score every item on exactly these levels, and report relevance_score as the chosen level divided by the highest level:
{relevance_levels}

Use only these inputs to complete the task defined by the system prompt.