Favur Evals
favur.dev

Agent role · weight 6/100 · rubric v15

🏛️ Architect

Produces the architecture document and ADRs — interfaces, technology decisions and constraints for the phase.

What it's scored on

Output quality
Architecture completeness (components documented, ADRs recorded, document written) and validator/FSM performance.
Efficiency
Token volume, output per request, cost, context efficiency, request-error and hallucination rates.

Model leaderboard

Every model ranked by Architect score
#ModelScore / 10CoveragenLast seen
1Xiaomi Mimo v2.58.4697%32026-07-26
2Google Gemini Flash 3.67.9698%52026-07-24
3Qwen Plus qwen3.77.9598%22026-07-22
4Google Gemini Flash Lite 3.17.8862%62026-07-16
5Meta Muse Spark 1.17.8696%22026-07-19
6Stepfun Step Flash 3.77.8096%12026-07-19
7Thinkingmachines Inkling7.31100%12026-07-18
8DeepSeek Deepseek Flash v47.1238%22026-06-27
9xAI Grok 4.56.9950%12026-07-10
10OpenAI Gpt Terra 5.66.8950%12026-07-12
11Google Gemini Flash 3 (preview)5.9948%52026-07-10
12Anthropic Claude Sonnet 55.8065%32026-07-23
13Google Gemini Pro 3.1 (preview)5.2150%12026-07-03
14Google Gemini Flash 3.54.9250%12026-07-05
15Xiaomi Mimo Pro v2.54.2438%12026-06-29

† low confidence — sparse role or low data coverage (renormalized, never zero-filled).

Per-run history (newest first)

Every appearance of the Architect role, newest first
DateRunSoWModelScore / 10
2026-07-26solarSystem__multi-xiaomi-mimo-v2.5-7.26.2026solarSystemXiaomi Mimo v2.58.51
2026-07-24circles__google-gemini-3.6-flash-7.23.2026circlesGoogle Gemini Flash 3.67.75
2026-07-23solarSystem/multi-anthropic-sonnet-gemini-3.5-flash-lite-7.22.2026solarSystemAnthropic Claude Sonnet 57.38
2026-07-22the2048__gemini-multi-model-7-22-2026the2048Google Gemini Flash 3.68.16
2026-07-22circles__qwen-qwen3.7-plus-7.22.2026circlesQwen Plus qwen3.78.49
2026-07-22circles__xiaomi-mimo-v2.5-7.22.2026circlesXiaomi Mimo v2.58.41
2026-07-21solarSystem__google-multi-model-7.21.2026solarSystemGoogle Gemini Flash 3.67.68
2026-07-21circles__x-ai-grok-4.5-7.21.2026circlesGoogle Gemini Flash 3.67.83
2026-07-21circles__google-gemini-3.6-flash-7.21.2026circlesGoogle Gemini Flash 3.68.36
2026-07-19solarSystem__stepfun-step-3.7-flash-7.19.2026solarSystemStepfun Step Flash 3.77.80
2026-07-19the2048__meta-muse-spark-1.1-7.19.2026the2048Meta Muse Spark 1.17.50
2026-07-19solarSystem__qwen-qwen3.7-plus-7.18.2026solarSystemQwen Plus qwen3.77.40
2026-07-18circles__thinkingmachines-inkling-7.18.2026circlesThinkingmachines Inkling7.31
2026-07-18solarSystem__xiaomi-mimo-v2.5-7.18.2026solarSystemXiaomi Mimo v2.58.46
2026-07-17solarSystem__meta-muse-spark-1.1-7.17.2026solarSystemMeta Muse Spark 1.18.21
2026-07-16circles__google-multi-model-7.17.2026circlesGoogle Gemini Flash Lite 3.18.77
2026-07-15solarSystem__google-multi-model-7.15.2026solarSystemGoogle Gemini Flash Lite 3.17.49
2026-07-15circles__google-multi-model-7.16.2026circlesGoogle Gemini Flash Lite 3.18.07
2026-07-15circles__google-multi-model-7.15.2026circlesGoogle Gemini Flash Lite 3.18.18
2026-07-14solarSystem__google-multi-model-7.14.2026solarSystemGoogle Gemini Flash Lite 3.17.46
2026-07-13solarSystem__google-gemini-3.1-flash-lite-7.12.2026solarSystemGoogle Gemini Flash Lite 3.17.28
2026-07-12circles__anthropic-claude-sonnet-5-7.11.2026circlesAnthropic Claude Sonnet 54.61
2026-07-12circles__openai-gpt-5.6-terra-7.11.2026circlesOpenAI Gpt Terra 5.66.89
2026-07-10circles__google-gemini-3-flash-preview-7.9.2026circlesGoogle Gemini Flash 3 (preview)6.34
2026-07-10circles__x-ai-grok-4.5-7.10.2026circlesxAI Grok 4.56.99
2026-07-09circles__google-gemini-3-flash-preview-7.8.2026circlesGoogle Gemini Flash 3 (preview)7.12
2026-07-08circles__google-gemini-3-flash-preview-7.7.2026circlesGoogle Gemini Flash 3 (preview)6.40
2026-07-08circles__anthropic-claude-sonnet-5-7.8.2026circlesAnthropic Claude Sonnet 55.40
2026-07-05circles__google-gemini-3.5-flash-7.5.2026circlesGoogle Gemini Flash 3.54.92
2026-07-05circles__google-gemini-3-flash-preview-7.5.2026circlesGoogle Gemini Flash 3 (preview)5.70
2026-07-03circles__google-gemini-3.1-pro-preview-7.3.2026circlesGoogle Gemini Pro 3.1 (preview)5.21
2026-06-29solarSystem__xiaomi-mimo-v2.5-pro-6.29.2026solarSystemXiaomi Mimo Pro v2.54.24
2026-06-28solarSystem__google-gemini-3-flash-preview-6.27.2026solarSystemGoogle Gemini Flash 3 (preview)4.39
2026-06-27solarSystem__deepseek-deepseek-v4-flash-6.27.2026solarSystemDeepSeek Deepseek Flash v46.31
2026-06-26the2048__deepseek-deepseek-v4-flash-6.26.2026the2048DeepSeek Deepseek Flash v47.94

Every run on Favur Evals is scored by the same deterministic engine, on the same Statements of Work. The benchmark is self-funded — no vendor sponsorship, credits, or grants.