Entity trail

DRACO

Source-backed findings, relationship evidence, citations, and briefing history from the public MindPattern archive.

Briefing refs

Findings

Edges

Sources

Corpus findings

2026-06-16 / hn-researcherOpenRouter Launches Fusion API — Model Panels at Half the CostOpenRouter's Fusion API (210 points, 81 comments) routes a prompt across multiple LLMs, uses a judge model to analyze outputs, then synthesizes a final answer, scoring 69% on DRACO and claiming frontier quality at half the price. Developer reaction split sharply: some called it a breakthrough compound model, others noted mixture-of-agents has existed since 2024 and that DRACO covers zero coding tasks. Consensus on HN: treat it as an escalation lane, not a default.
2026-06-15 / news-researcherOpenRouter Launches 'Fusion,' a Compound-Model API That Beats Frontier Models at Roughly Half the CostOpenRouter released Fusion, a compound API that fans each prompt out to a panel of models, synthesizes their answers, and returns one response. On Perplexity's DRACO deep-research benchmark (100 tasks across 10 domains), Fable 5 + GPT-5.5 fused scored 69.0% versus 65.3% for Fable 5 alone, and a budget panel of Gemini 3 Flash + Kimi K2.6 + DeepSeek V4 Pro beat solo GPT-5.5 and Opus 4.8 at about half Fable 5's price — though calls run 2-3x longer.
2026-03-31 / skill-finderMicrosoft Copilot Critique: GPT Drafts, Claude Reviews — 13.8% DRACO Benchmark Improvement via Multi-Model Orchestration in Enterprise ResearchMicrosoft announced Critique for Microsoft 365 Copilot Researcher on March 30: GPT drafts a research response, then Anthropic's Claude reviews it for accuracy, completeness, and citation quality before delivery. Microsoft reports a 13.8-point improvement on the DRACO benchmark. Council mode enables side-by-side model comparisons, with plans for bidirectional critique. This is the clearest signal yet that enterprise AI is moving from single-model to multi-model orchestration as a competitive advantage.

Graph relationships

BENCHMARKED AGAINST
Fusion
Fusion scored 69.0% on DRACO benchmark vs 65.3% for Fable 5 alone.
Source finding
BENCHMARKED AGAINST
Fusion -> DRACO
Fusion scored 69.0% on DRACO benchmark vs 65.3% for Fable 5 alone.
Source finding

Source trail

Codex Knowledge Base writeup OpenRouter OpenRouter Critique 13.8% improvement on the DRACO benchmark Composio open-sourced Agent Orchestrator

Graph sources

entity graphfindings textkg edgeskg entitiesnewsletter issues

Fetching from the wire…