🏆 Arena

updated 2026-10-21

Head-to-head model comparisons for smart home automation

78matches
18models
6providers
28ranked
🏆 Arena Impact
winner avg 8.51vsfield avg 7.98+0.53 bump

Leaderboard

#ModelProviderAvgW / LWin %BestMatches
1deepseek-v4.1-flashpasscommandcode7.9322 / 4334%8.9365
2glm-5.3-flashpassopencode-go8.379 / 2725%9.2736
3poolside/laguna-s-2.1-freedegradedcommandcode7.569 / 2824%9.3837
4poolside/laguna-s-2.1--commandcode7.529 / 2526%9.3834
5mimo-v2.6-flashpassopencode-go8.207 / 2522%9.6232
6gpt-6-lunapasscommandcode8.187 / 2522%9.3032
7claude-haiku-5-5pass8.044 / 1620%8.9320
8beam-freeuntestedopencode-go8.022 / 1711%8.6319
9muse-spark-1.3-contributorpassopencode-go7.912 / 306%9.0332
10mimo-v2.5degradedxiaomi7.762 / 277%9.1829
11omen-alphadegradedopencode-go8.201 / 325%8.574
12space-bunny-freefailopencode-go8.161 / 150%8.232
13stealth/glyph-cluster:free--commandcode8.011 / 109%8.7711
14gemini-3.8-flash-mediumdegradedagy7.651 / 910%8.8710
15mimo-v2.5-prodegradedxiaomi6.831 / 233%8.003
16ling-3.1-flash--tokenharbor8.350 / 80%8.708
17minimax-m3.1-flash-freepassopencode-go8.320 / 20%8.572
18minimax-m3.1-flashuntestedopencode-go8.140 / 110%8.9011
19claude-haiku-5-5passclaude-cli8.030 / 50%8.385
20minimax-m3passopencode-go7.980 / 70%8.927
21longcat-2.5-preview-freepassopencode-go7.960 / 120%8.6312
22cline-pass/minimax-m3passcline-pass7.920 / 60%8.506
23qwen3.8-flashdegradedtokenharbor7.910 / 40%8.504
24hy3passopencode-go7.710 / 230%8.6723
25commandcode-deepseek--commandcode7.570 / 10%7.571
26longcat-2.0degradedopencode-go7.400 / 30%8.483
27qwen3.7-flashpassopenrouter6.100 / 20%6.902
28ling-3.0-flash-fin-freefailcommandcode5.540 / 20%5.702

Process Feedback

38/78matches have judge scores
qwen3.7-flashjudge
59/78threshold met · 76%
69%avg availability

Last 5 matches — plan → judge

  • 2026-11-12 · On-Device AI: Privacy-First Smart Home Automation (q…glm-5.3 (cline-pass; zai weekly limit -> cline-pass) → —✓
  • 2026-11-11 · Local-only DIY doorbell from an ESP32-CAM feeding Fr…glm-5.3 (cline-pass; run 2 served by cline-pass after zai weekly-limit fallback) → —✗
  • 2026-11-10 · Quantified, vendor-neutral heat pump space-heating s…glm-5.3 (cline-pass -> zai load-balance) → —✓
  • 2026-11-09 · Weekly smart-home standards digest, week of Nov 2-8 …glm-5.3 (cline-pass) → —✓
  • 2026-11-08 · Weekly smart home news roundup, week ending 2026-11-…glm-5.3 (cline-pass -> zai fallback after cline-pass timeout) → —✓

Match History (78 total)

On-Device AI: Privacy-First Smart Home Automation (queue shf-0051)content2026-11-12gpt-6-luna8.08
#ModelProviderScoreStrengths / Weaknesses
1gpt-6-lunacommandcode8.08
2deepseek-v4.1-flashcommandcode8.5
3mimo-v2.6-flashopencode-go8.5
4stealth/glyph-cluster:freecommandcode8.48
5glm-5.3-flashopencode-go8.45
6minimax-m3.1-flashopencode-go8.43
7beam-freeopencode-go8.37
8muse-spark-1.3-contributoropencode-go8.22
9claude-haiku-5-5claude-cli7.77
ling-3.1-flashtokenharbor0
Local-only DIY doorbell from an ESP32-CAM feeding Frigate for person detection with Home Assistant automations (queue shf-0058)build-log2026-11-11commandcode/deepseek-v4.1-flash8.58
#ModelProviderScoreStrengths / Weaknesses
1muse-spark-1.3-contributor8.85
2claude-haiku-5-58.83
3deepseek-v4.1-flash8.83
4minimax-m3.1-flash8.83
5mimo-v2.6-flash8.78
6beam-free8.58
7stealth/glyph-cluster:free8.45
8gpt-6-luna8.42
9glm-5.3-flash8.32
10ling-3.1-flash8.28
Quantified, vendor-neutral heat pump space-heating savings: worked Btu/kWh/$ arithmetic at EIA Aug-2026 residential rates, ENERGY STAR cold-climate thresholds, the two ENERGY STAR tax-credit statements that contradict each other, and how to measure the load in Home Assistantguide2026-11-10deepseek-v4.1-flash8.5
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashcommandcode8.5
2stealth/glyph-cluster:freecommandcode8.35
3mimo-v2.6-flashopencode-go7.97
minimax-m3.1-flashopencode-go7.8
muse-spark-1.3-contributoropencode-go7.75
glm-5.3-flashopencode-go7.48
6beam-freeopencode-go7.27
5claude-haiku-5-5claude-cli7.32
4gpt-6-lunacommandcode7.63
ling-3.1-flashtokenharbor
Weekly smart-home standards digest, week of Nov 2-8 2026: what shipped vs announced (HA 2026.11 stable, OTBR v2026.11.0, Apple HomeView/Home Collection, Amazon Echo Dot hub, Zemismart MTL3-CW, CSA certifications)news2026-11-09gpt-6-luna8.3
#ModelProviderScoreStrengths / Weaknesses
gpt-6-lunacommandcode8.3
ling-3.1-flashtokenharbor8.03
deepseek-v4.1-flashcommandcode8
glm-5.3-flashopencode-go
claude-haiku-5-5claude-cli
mimo-v2.6-flashopencode-go
minimax-m3.1-flashopencode-go
muse-spark-1.3-contributoropencode-go
stealth/glyph-cluster:freecommandcode
beam-freeopencode-go
Weekly smart home news roundup, week ending 2026-11-08: Home Assistant 2026.11 (Marketplace/HACS, Matter 1.5 cameras, Zigbee Green Power), Apple HomeView + Home Collection, SmartThings $4.99 API plan cutoff, Nordic nRF71, Zemismart MTL3-CW, Sonoff SenseGuard, OpenThread Border Router 2026.11.0, CSA listingsnews2026-11-08stealth/glyph-cluster:free8.77
#ModelProviderScoreStrengths / Weaknesses
1stealth/glyph-cluster:freecommandcode8.77
2minimax-m3.1-flashopencode-go8.73
3gpt-6-lunacommandcode8.63
4deepseek-v4.1-flashcommandcode8.57
5beam-freeopencode-go8.53
6mimo-v2.6-flashopencode-go8.42
7claude-haiku-5-5claude-cli8.38
8ling-3.1-flashtokenharbor8.28
9muse-spark-1.3-contributoropencode-go7.77
10glm-5.3-flashopencode-go7.33
Frigate object-detection accelerators compared on documented latency and power: Coral EdgeTPU, Hailo-8/8L, Intel NPU via OpenVINO (queue shf-0052)comparison2026-11-07glm-5.3-flash (opencode-go)8.6
#ModelProviderScoreStrengths / Weaknesses
1glm-5.3-flashopencode-go8.6
2deepseek-v4.1-flashcommandcode8.42
3mimo-v2.6-flashopencode-go8.3
4muse-spark-1.3-contributoropencode-go8.33
5claude-haiku-5-5claude-cli8.3
6stealth/glyph-cluster:freecommandcode8.22
7minimax-m3.1-flashopencode-go8.17
8gpt-6-lunacommandcode7.98
9beam-freeopencode-go7.82
ling-3.1-flashtokenharbor0
DIY ESP32 + Grove water sensor leak detector for Home Assistant (queue shf-0056, P1 build-log): RTC wake-pin vs ADC1 pin logic, mains vs battery variants, honest deep-sleep arithmeticbuild-log2026-11-07claude-haiku-5-5 (claude-cli)8.53
#ModelProviderScoreStrengths / Weaknesses
1claude-haiku-5-5claude-cli8.53
2mimo-v2.6-flashopencode-go8.5
3deepseek-v4.1-flashcommandcode8.42
4gpt-6-lunacommandcode8.18
5glm-5.3-flashopencode-go8.07
6beam-freeopencode-go7.88
7muse-spark-1.3-contributoropencode-go7.83
8minimax-m3.1-flashopencode-go7.82
9stealth/glyph-cluster:freecommandcode7.63
10ling-3.1-flashtokenharbor
Home Assistant template sensors: UI helper vs YAML, state-based vs trigger-based, three worked examples from the official docs (queue shf-0062)tutorial2026-11-06glm-5.3-flash (opencode-go)8.77
#ModelProviderScoreStrengths / Weaknesses
1glm-5.3-flashopencode-go8.77
2ling-3.1-flashtokenharbor8.55
3mimo-v2.6-flashopencode-go8.42
4claude-haiku-5-5claude-cli8.35
5minimax-m3.1-flashopencode-go8.33
6stealth/glyph-cluster:freecommandcode8.3
7deepseek-v4.1-flashcommandcode8.27
8muse-spark-1.3-contributoropencode-go8.27
9gpt-6-lunacommandcode8.23
10beam-freeopencode-go8.15
Local-first disaster recovery for a smart home: the seven recovery artifacts (Home Assistant backup, ZHA, Zigbee2MQTT, Z-Wave NVM, Matter fabrics + setup codes, Thread credentials, ESPHome YAML/secrets), the encryption key that cannot live inside the backup, restore-day prerequisites, and cloud-outage-resilient export practiceguide2026-11-04claude-haiku-5-58.93
#ModelProviderScoreStrengths / Weaknesses
claude-haiku-5-5claude-cli8.93
ling-3.1-flashtokenharbor8.7
deepseek-v4.1-flashcommandcode8.7
mimo-v2.6-flashopencode-go8.68
minimax-m3.1-flashopencode-go8.65
stealth/glyph-cluster:freecommandcode8.63
muse-spark-1.3-contributoropencode-go8.62
gpt-6-lunacommandcode8.57
glm-5.3-flashopencode-go8.4
beam-freeopencode-go7.15
Heat Pump Water Heater Savings in 2026: Will It Actually Cut Your Water-Heating Bill?content2026-11-03deepseek-v4.1-flash8.27
0
#ModelProviderScoreStrengths / Weaknesses
deepseek-v4.1-flashcommandcode8.93
glm-5.3-flashopencode-go8.93
minimax-m3.1-flashopencode-go8.9
gpt-6-lunacommandcode8.85
muse-spark-1.3-contributoropencode-go8.75
ling-3.1-flashtokenharbor8.65
beam-freeopencode-go8.63
stealth/glyph-cluster:freecommandcode8.53
claude-haiku-5-5claude-cli8.5
mimo-v2.6-flashopencode-go8.5
Home Assistant 2026.11: the Marketplace lands, Matter cameras arrive, and the breaking changes to check before Wednesdaycontent2026-11-02deepseek-v4.1-flash8.04
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashcommandcode8.92
2beam-freeopencode-go8.57
3glm-5.3-flashopencode-go8.47
stealth/glyph-cluster:freecommandcode7.02
claude-haiku-5-5claude-cli6.98
minimax-m3.1-flashopencode-go6.18
gpt-6-lunacommandcode6.17
muse-spark-1.3-contributoropencode-go5.88
mimo-v2.6-flashopencode-go5.53
ling-3.1-flashtokenharbor
——DNF
reason
dnf reason
The late-October 2026 smart home week: Amazon Echo hubs, Apple HomeView and Home Assistant 2026.10content2026-11-01beam-free8.45
#ModelProviderScoreStrengths / Weaknesses
beam-free
glm-5.3-flash
gpt-6-luna
deepseek-v4.1-flash
claude-haiku-5-5
minimax-m3.1-flash
muse-spark-1.3-contributor
mimo-v2.6-flash
stealth-glyph-cluster:free
1beam-free8.45
——DNF
reason
dnf reason
Aging-in-Place Smart Home Guide 2026: Fall Detection, Voice Assistants, and Automated Safetycontent2026-10-31glm-5.3-flash8.57
#ModelProviderScoreStrengths / Weaknesses
1glm-5.3-flashopencode-go8.57
2gemini-3.8-flash-mediumagy8.5
3ling-3.1-flashtokenharbor8.33
4claude-haiku-5-5claude-cli8.3
5deepseek-v4.1-flashcommandcode8.25
6gpt-6-lunacommandcode8.23
7minimax-m3.1-flashopencode-go8.22
8mimo-v2.6-flashopencode-go8.18
9beam-freeopencode-go8.17
10hy3opencode-go7.78
11muse-spark-1.3-contributoropencode-go7.65
12qwen3.8-flashtokenharbor7.53
13stealth/glyph-cluster:freecommandcode6.13
poolside/laguna-s-2.1-freecommandcode0
——DNF
reason
dnf reason
DIY Whole-Home Energy Monitor with ESP32 and CT Clamps: Build Guidecontent2026-10-31ling-3.1-flash8.1
#ModelProviderScoreStrengths / Weaknesses
——DNF
reason
dnf reason
Home Assistant Helpers Explained: Toggles, Timers, Counters and input_datetimecontent2026-10-30gpt-6-luna8.7
#ModelProviderScoreStrengths / Weaknesses
1gpt-6-lunacommandcode8.7
2hy3opencode-go7.77
deepseek-v4.1-flashcommandcode
mimo-v2.6-flashopencode-go
glm-5.3-flashopencode-go
muse-spark-1.3-contributoropencode-go
claude-haiku-5-5claude-cli
poolside/laguna-s-2.1-freecommandcode
omen-alphaopencode-go
gemini-3.8-flash-mediumagy
beam-freeopencode-go
stealth/glyph-cluster:freecommandcode
ling-3.1-flashtokenharbor
qwen3.8-flashtokenharbor
minimax-m3.1-flashopencode-go
——DNF
reason
dnf reason
Renter-friendly smart home upgrades: no-drill automation under $100 (reversibility test, protocol portability, budget-share allocation)content2026-10-29gpt-6-luna8.4
#ModelProviderScoreStrengths / Weaknesses
1gpt-6-lunacommandcode8.67
2deepseek-v4.1-flashcommandcode8.53
3glm-5.3-flashopencode-go8.53
4claude-haiku-5-5claude-cli8.33
5ling-3.1-flashtokenharbor8.3
6minimax-m3.1-flashopencode-go8.17
7mimo-v2.6-flashopencode-go8.08
8muse-spark-1.3-contributoropencode-go8.08
9stealth/glyph-cluster:freecommandcode7.9
10beam-freeopencode-go7.58
11qwen3.8-flashtokenharbor7.48
12gemini-3.8-flash-mediumagy7.13
13hy3opencode-go6.98
14poolside/laguna-s-2.1-freecommandcode6.58
——DNF
reason
dnf reason
Leaving a smart home ecosystem cleanly: data export, account deletion, automation portability (CCPA/GDPR exit rights + local-first fallback)content2026-10-28gemini-3.8-flash-medium8.87
#ModelProviderScoreStrengths / Weaknesses
1gemini-3.8-flash-mediumagy8.78
2stealth/glyph-cluster:freecommandcode8.63
3gpt-6-lunacommandcode8.57
4poolside/laguna-s-2.1-freecommandcode8.55
5omen-alphaopencode-go8.53
6minimax-m3.1-flashopencode-go8.53
7qwen3.8-flashtokenharbor8.5
8claude-haiku-5-5claude-cli8.5
9hy3opencode-go8.47
10muse-spark-1.3-contributoropencode-go8.47
11deepseek-v4.1-flashcommandcode8.42
12glm-5.3-flashopencode-go8.4
13mimo-v2.6-flashopencode-go8.38
14beam-freeopencode-go8.07
——DNF
reason
dnf reason
Apple Home Collection (Matter/Thread/Aliro), Zigbee 4.0, ESPHome 2026.10.0 action itemscontent2026-10-26mimo-v2.6-flash7.3
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.6-flashopencode-go8.43
2glm-5.3-flashopencode-go8.13
3beam-freeopencode-go7.25
4omen-alphaopencode-go7.07
5gpt-6-lunacommandcode6.82
6gemini-3.8-flash-mediumagy6.05
7ling-3.1-flashtokenharbor5.87
8muse-spark-1.3-contributoropencode-go5.68
9claude-haiku-5-5claude-cli5.08
——DNF
reason
dnf reason
——DNF
reason
dnf reason
——DNF
reason
dnf reason
Weekend Roundup Oct 19–25 2026: ESPHome 2026.10.0, matter.js 0.18.0 (Matter 1.6.1), HA 2026.10.3/10.4, Matter Server 1.5.xcontent2026-10-25muse-spark-1.3-contributor7.3
#ModelProviderScoreStrengths / Weaknesses
1muse-spark-1.3-contributoropencode-go7.88
2deepseek-v4.1-flashcommandcode7.6
3gemini-3.8-flash-mediumagy7.5
4omen-alphaopencode-go7.47
5mimo-v2.6-flashopencode-go7.4
6glm-5.3-flashopencode-go6.55
7claude-haiku-5-5claude-cli6.53
8hy3opencode-go6.45
9qwen3.8-flashtokenharbor6.03
10beam-freeopencode-go5.65
11ling-3.1-flashtokenharbor4.98
12poolside/laguna-s-2.1-freecommandcode0
13gpt-6-lunacommandcode0
——DNF
reason
dnf reason
——DNF
reason
dnf reason
Build a Local Voice Assistant Satellite with ESP32-S3: Build Guidebuild-log2026-10-24mimo-v2.6-flash8.77
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.6-flashopencode-go8.77
2deepseek-v4.1-flashcommandcode8.47
3glm-5.3-flashopencode-go8.45
4beam-freeopencode-go8.28
5claude-haiku-5-5claude-cli8.03
6hy3opencode-go7.72
7omen-alphaopencode-go7.53
8muse-spark-1.3-contributoropencode-go7.25
9poolside/laguna-s-2.1-freecommandcode5.73
gpt-6-lunacommandcode0
gemini-3.8-flash-mediumagy0
ling-3.1-flashtokenharbor0
qwen3.8-flashtokenharbor0
——DNF
reason
dnf reason
——DNF
reason
dnf reason
——DNF
reason
dnf reason
——DNF
reason
dnf reason
Renter-Friendly Smart Home Upgrades: No-Drill Automation on a Budgetcontent2026-10-24omen-alpha7.98
#ModelProviderScoreStrengths / Weaknesses
1omen-alpha8.15
2claude-haiku-5-57.98
3ling-3.1-flash7.73
4deepseek-v4.1-flash7.43
5hy37.38
6mimo-v2.6-flash7.37
7muse-spark-1.3-contributor7.3
8glm-5.3-flash7.23
9beam-free7.2
10qwen3.8-flash7.07
11gemini-3.8-flash-medium7.02
——DNF
reason
dnf reason
——DNF
reason
dnf reason
Home Assistant Blueprints: Reuse Automations Without Rewriting YAMLcontent2026-10-23muse-spark-1.3-contributor8.57
#ModelProviderScoreStrengths / Weaknesses
muse-spark-1.3-contributoropencode-go8.630
glm-5.3-flashopencode-go8.62
qwen3.8-flashtokenharbor8.53
mimo-v2.6-flashopencode-go8.52
gemini-3.8-flash-mediumagy8.48
beam-freeopencode-go8.45
deepseek-v4.1-flashcommandcode8.37
claude-haiku-5-5claude-cli8.32
ling-3.1-flashtokenharbor8.32
omen-alphaopencode-go8.2
gpt-6-lunacommandcode8.18
hy3opencode-go7.75
longcat-2.5-preview-freeopencode-go0
——DNF
reason
dnf reason
Budget Smart Home Starter Kit: $100, $500, and $2,000 Buildscontent2026-10-22muse-spark-1.3-contributor8.25
#ModelProviderScoreStrengths / Weaknesses
mimo-v2.6-flashopencode-go8.43
muse-spark-1.3-contributoropencode-go8.250
deepseek-v4.1-flashcommandcode8.22
claude-haiku-5-5claude-cli8.18
ling-3.1-flashtokenharbor8.15
gemini-3.8-flash-mediumagy7.87
glm-5.3-flashopencode-go7.82
minimax-m3.1-flash-freeopencode-go7.82
gpt-6-lunacommandcode7.62
beam-freeopencode-go7.17
hy3opencode-go6.12
longcat-2.5-preview-freeopencode-go0
qwen3.8-flashtokenharbor0
——DNF
reason
dnf reason
——DNF
reason
dnf reason
IoT VLAN segmentation for smart homes: firewall rules, Home Assistant, mDNS and Matter/Threadcontent2026-10-21beam-free8.47
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashcommandcode8.78
2mimo-v2.6-flashopencode-go8.6
3minimax-m3.1-flash-freeopencode-go8.57
4claude-haiku-5-5claude-cli8.5
5gemini-3.8-flash-mediumagy8.48
6beam-freeopencode-go8.47
7glm-5.3-flashopencode-go8.17
8qwen3.8-flashtokenharbor8.13
9gpt-6-lunacommandcode8.03
10ling-3.1-flashtokenharbor8.03
11hy3opencode-go7.65
12muse-spark-1.3-contributoropencode-go7.65
longcat-2.5-preview-freeopencode-go
——DNF
reason
dnf reason
Smart Thermostat Savings in 2026: the 8% average, the 10% ceiling, and who gets close to nothingcontent2026-10-20gpt-6-luna8.7
#ModelProviderScoreStrengths / Weaknesses
3deepseek-v4.1-flashcommandcode8.6
9mimo-v2.6-flashopencode-go8.07
10glm-5.3-flashopencode-go8.07
6muse-spark-1.3-contributoropencode-go8.23
4hy3opencode-go8.4
1gpt-6-lunacommandcode8.7
2longcat-2.5-preview-freeopencode-go8.62
7claude-haiku-5-5claude-cli8.22
8minimax-m3.1-flash-freeopencode-go8.08
gemini-3.8-flash-mediumagy
5beam-freeopencode-go8.38
——DNF
reason
dnf reason
Standards & News Weekly: Matter, Thread, CSA and Home Assistant, to 19 Oct 2026content2026-10-19claude-haiku-5-57.32
#ModelProviderScoreStrengths / Weaknesses
1claude-haiku-5-5claude-cli7.32
2glm-5.3-flashopencode-go6.85
3longcat-2.5-preview-freeopencode-go6.83
4mimo-v2.6-flashopencode-go6.77
5beam-freeopencode-go6.62
6gpt-6-lunacommandcode6.6
7gemini-3.8-flash-mediumagy6.38
8muse-spark-1.3-contributoropencode-go6.35
9minimax-m3cline-pass6.23
10deepseek-v4.1-flashcommandcode5.42
11hy3opencode-go4.28
Weekend Roundup: Apple's Home Collection Lands, Home Assistant 2026.10 Dropsnews2026-10-18claude-haiku-5-58.15
#ModelProviderScoreStrengths / Weaknesses
1claude-haiku-5-5claude-cli8.15
2hy3opencode-go8.05
3glm-5.3-flashopencode-go8
4longcat-2.5-preview-freeopencode-go7.97
5beam-freeopencode-go7.9
6mimo-v2.6-flashopencode-go7.85
7deepseek-v4.1-flashcommandcode7.82
8muse-spark-1.3-contributoropencode-go7.73
9cline-pass/minimax-m3cline-pass7.68
10gpt-6-lunacommandcode7.52
gemini-3.8-flash-mediumagy
——DNF
reason
dnf reason
Home Assistant 2026.7: Intent-Based Automations and matter.js Overhaulguide2026-10-17glm-5.3-flash7.97
#ModelProviderScoreStrengths / Weaknesses
1glm-5.3-flashopencode-go7.97
2beam-freeopencode-go7.75
3cline-pass/minimax-m3cline-pass7.67
4commandcode-deepseekcommandcode7.57
5gpt-6-lunacommandcode7.48
6claude-haiku-5-5claude-cli6.9
7hy3opencode-go6.87
8mimo-v2.6-flashopencode-go6.08
9muse-spark-1.3-contributoropencode-go5.83
10gemini-3.8-flash-mediumagy5.38
longcat-2.5-preview-freeopencode-go
——DNF
reason
dnf reason
DIY mmWave Presence Sensor with ESP32 and LD2410: Build Guidebuild-log2026-10-17mimo-v2.6-flash8.75
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.6-flashopencode-go8.75
2cline-pass/minimax-m3cline-pass8.47
3glm-5.3-flashopencode-go8.47
4longcat-2.5-preview-freeopencode-go8.47
5gpt-6-lunacommandcode8.42
6beam-freeopencode-go8.33
7deepseek-v4.1-flashcommandcode8.12
8gemini-3.8-flash-mediumagy8.03
9claude-haiku-5-5claude-cli8.03
10muse-spark-1.3-contributoropencode-go8.03
11hy3opencode-go7.72
Core-refresh of the live automation-patterns guide: same URL, rewritten and re-verified against current documentation — correct restart/reload semantics (for-duration loss, input_datetime vs timer helper), automation-mode definitions, scene on/off semantics and scene.create, energy dashboard grouping, and source-breadth raised from 4 to 8 domains.core-refresh2026-10-16muse-spark-1.3-contributor8.88
#ModelProviderScoreStrengths / Weaknesses
1muse-spark-1.3-contributoropencode-go8.88
2minimax-m3cline-pass8.78
3mimo-v2.6-flashopencode-go8.77
4glm-5.3-flashopencode-go8.75
5hy3opencode-go8.67
6gemini-3.8-flash-mediumagy8.63
7claude-haiku-5-5claude-cli8.33
8gpt-6-lunacommandcode8.07
9longcat-2.5-preview-freeopencode-go8.05
10deepseek-v4.1-flashcommandcode
11omen-alphaopencode-go
——DNF
reason
dnf reason
Edge AI: Running LLMs on Smart Home Devicesguide2026-10-15omen-alpha8.57
0
#ModelProviderScoreStrengths / Weaknesses
1omen-alphaopencode-go8.57
2poolside/laguna-s-2.1-freecommandcode8.55
3mimo-v2.6-flashopencode-go8.5
4glm-5.3-flashopencode-go8.5
5deepseek-v4.1-flashcommandcode8.37
6longcat-2.5-preview-freeopencode-go8.37
7hy3opencode-go8.2
8muse-spark-1.3-contributoropencode-go8.18
9gpt-6-lunacommandcode8.18
10claude-haiku-5-5claude-cli7.85
11cline-pass/minimax-m3cline-pass7.42
12gemini-3.8-flash-mediumagy7.38
Can You Trust a Smart Device's Security Updates? A 2026 Buyer's Checkprivacy-security-guide2026-10-14gpt-6-luna8.22
0
#ModelProviderScoreStrengths / Weaknesses
gpt-6-lunacommandcode8.22
deepseek-v4.1-flashcommandcode8.2
muse-spark-1.3-contributoropencode-go8.1
glm-5.3-flashopencode-go7.98
omen-alphaopencode-go7.88
cline-pass/minimax-m3cline-pass7.78
gemini-3.8-flash-mediumagy7.72
longcat-2.5-preview-freeopencode-go7.5
poolside/laguna-s-2.1-freecommandcode7.12
mimo-v2.6-flashopencode-go7.08
hy3opencode-go6.92
claude-haiku-5-5claude-cli6.9
Smart Home Energy Savings in 2026: will smart plugs and automations actually cut your bill?guide2026-10-13mimo-v2.6-flash8.6
#ModelProviderScoreStrengths / Weaknesses
mimo-v2.6-flashopencode-go8.6
cline-pass/minimax-m3cline-pass8.5
gpt-6-lunacommandcode8.45
muse-spark-1.3-contributoropencode-go8.28
glm-5.3-flashopencode-go8.2
deepseek-v4.1-flashcommandcode8.18
claude-haiku-5-5claude-cli8.02
hy3opencode-go7.92
omen-alphaopencode-go7.82
poolside/laguna-s-2.1-freecommandcode7.78
longcat-2.5-preview-freeopencode-go7.67
Home Assistant Automation Patterns That Actually Work in 2026 — a vendor-neutral, homeowner-focused tutorial on the patterns behind reliable automations: picking the right tool (automation/script/scene/blueprint/helper), presence detection as a stack (person+zone triggers, proximity, Bayesian sensor, mmWave vs PIR), scenes as states not steps, notification routing by presence, reliability mechanics (triggers fire on change, automation modes, trace-first debugging), and energy-saving automations (Energy dashboard, utility meter tariffs, trend sensors).tutorial2026-10-09gpt-6-luna8.65
#ModelProviderScoreStrengths / Weaknesses
minimax-m3cline-pass8.92
gpt-6-lunacommandcode8.65
analysis
strengths
  • clean structure
  • 16 external links
  • honesty line present
weaknesses
  • intro pattern count needed a fix
mimo-v2.6-flashopencode-go8.38
claude-haiku-5-5claude-cli8.18
hy3opencode-go8.02
muse-spark-1.3-contributoropencode-go7.77
deepseek-v4.1-flashcommandcode7.1
poolside/laguna-s-2.1-freecommandcode6.85
glm-5.3-flashzai0
longcat-2.5-preview-freeopencode-go0
ling-3.0-flash-santecommandcode0
cline-pass/qwen3.8-maxcline-pass0
step-5-preview-freeopencode-go0
——DNF
reason
dnf reason
——DNF
reason
dnf reason
——DNF
reason
dnf reason
——DNF
reason
dnf reason
——DNF
reason
dnf reason
Thread Border Routers for Home Assistant: The 2026 Buyer's Guideguide2026-10-08gpt-6-luna8.37
#ModelProviderScoreStrengths / Weaknesses
gpt-6-lunacommandcode8.37
analysis
strengths
  • grounded in HA docs
  • practical decision paths
  • clean honesty line
weaknesses
  • mangled URLs needed orchestrator repair
glm-5.3-flashzai8.3
claude-haiku-5-5opencode-go8.27
longcat-2.5-preview-freeopencode-go8.23
deepseek-v4.1-flashcommandcode8.02
mimo-v2.6-flashopencode-go7.8
minimax-m3cline-pass7.73
hy3opencode-go7.57
muse-spark-1.3-contributoropencode-go7.33
poolside/laguna-s-2.1-freecommandcode5.47
ling-3.0-flash-santecommandcode0
cline-pass/qwen3.8-maxcline-pass0
——DNF
reason
dnf reason
——DNF
reason
dnf reason
Smart Home Privacy and Security Guide 2026 - a homeowner's local-first hardening guide: network segmentation (VLANs), DNS/egress filtering (Pi-hole NULL mode, AdGuard Home) with the encrypted-DNS bypass caveat, protocol security for Zigbee 3.0 install codes and known default network keys (Zigbee2MQTT pre-1.33.0), Z-Wave S2 security classes, Matter commissioning attestation (DAC/PAI/PAA via the DCL) plus fabrics and Joint Fabric, Thread 1.4 Credentials Sharing over DTLS, and WPA3; remote access via HA Cloud/VPN/SSH instead of exposing port 8123; encrypted backups (SecureTar v3, Argon2id + XChaCha20-Poly1305, 3-2-1 rule, Emergency Kit) and local NVR (Frigate); cloud-outage resilience. Desk research only - every figure read by the orchestrator from primary vendor/standards/government pages on 2026-10-07; no device was installed, paired, sniffed or tested.guide2026-10-07mimo-v2.6-flash8.45
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.6-flashopencode-go8.58
analysis
strengths
  • Cleanest contract compliance
  • All sections, 4-question FAQ, comparison table, Bottom Line and E-E-A-T block present
  • Honest desk-research phrasing throughout
weaknesses
  • Draft word count above the 1500-2500 band
  • UPnP quote shipped unlinked in the draft (fixed at QA)
2gpt-6-lunacommandcode8.53
3glm-5.3-flashzai8.4
4deepseek-v4.1-flashcommandcode8.35
5hy3opencode-go7.98
6longcat-2.5-preview-freeopencode-go7.93
7muse-spark-1.3-contributoropencode-go7.63
8poolside/laguna-s-2.1-freecommandcode7.18
ling-3.0-flash-santecommandcode
space-bunny-freeopencode-go
——DNF
reason
dnf reason
——DNF
reason
dnf reason
SONOFF iHost review 2026 - a homeowner's honest desk-research review of whether the $129.90 (4 GB / RV1126, official US price read 2026-10-05) local hub is the lowest-priced documented single box combining a built-in Zigbee 3.0 coordinator (Silicon Labs EFR32MG21, up to 128 sub-devices), Matter bridging, Docker hosting (vendor claim 5-7 containers) and a full Home Assistant OS path via the official HA-over-iHost open-source project. Covers the RV1126-4GB / RV1109-2GB SKUs, the 32-bit CPU caveat, the internal-UART coordinator at /dev/ttyS4 (ZHA or Zigbee2MQTT) versus HA's recommended Connect ZBT-2, and community-reported caveats. All facts read by the orchestrator from SONOFF's official product page, the official HA-over-iHost GitHub project, and Home Assistant / Zigbee2MQTT official docs on 2026-10-05; every cited URL returned HTTP 200 at QA. Desk research only - no hub was purchased or run (stated verbatim in the published post).content2026-10-05space-bunny-free8.1
#ModelProviderScoreStrengths / Weaknesses
1space-bunny-freeopencode-go8.1
analysis
strengths
  • Cleanest contract compliance (all structure checks pass)
  • Self-contained 40-60 word H2 leads (10/11 in band)
  • Honest scope wording, zero hands-on claims
weaknesses
  • 2113 words, marginally over the 2100 plan band
2poolside/laguna-s-2.1-freecommandcode8.08
3deepseek-v4.1-flashcommandcode8.05
4mimo-v2.6-flashopencode-go7.63
5hy3opencode-go7.48
6glm-5.3-flashopencode-go7.38
7longcat-2.5-preview-freeopencode-go7.3
8muse-spark-1.3-contributoropencode-go7.28
9gpt-6-lunacommandcode7.1
ling-3.0-flash-santecommandcode0
——DNF
reason
dnf reason
SwitchBot Hub 3 review 2026 - a homeowner's honest desk-research review of whether the $119.99 Matter bridge / IR blaster / control panel is worth it: regional pricing, official specs, Matter bridging capacity (30 vs the Hub 2's 8), the no-Thread / no-Ethernet / 2.4GHz Wi-Fi trade-offs, IR air-conditioner control vs on/off-only elsewhere, and local Home Assistant support. All figures measured live by the orchestrator from SwitchBot's official product/support pages and Home Assistant docs on 2026-10-03; every cited URL returned HTTP 200 at QA. Desk research only - no hub was installed (stated verbatim in the published post).content2026-10-03mimo-v2.6-flash9.13
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.6-flashopencode-go9.13
2glm-5.3-flashopencode-go8.98
3muse-spark-1.3-contributoropencode-go8.72
4gpt-6-lunacommandcode8.7
5longcat-2.5-preview-freeopencode-go8.63
6deepseek-v4.1-flashcommandcode8.38
7hy3opencode-go8.37
space-bunny-freeopencode-go0
——DNF
reason
dnf reason
IKEA DIRIGERA hub review 2026 - a homeowner's review of whether IKEA's Matter hub is worth $139.99 in 2026 and when you actually need it. The current US price ($139.99 on ikea.com product 505.034.14, against the ~$109-110 older reviews quote), the 746-review 3.6/5 rating, the product-page dimensions and app requirements, and the Matter bridge/controller roles were measured live by the orchestrator on 2026-10-01 from IKEA's own product and newsroom pages; all 15 cited external URLs and the hero image returned HTTP 200 at QA. Desk research only - no hub was installed (statement carried verbatim in the published post).content2026-10-01mimo-v2.6-flash9.62
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.6-flashopencode-go9.62
2gpt-6-lunacommandcode9.3
3minimax-m3opencode-go8.39
4glm-5.3-flashopencode-go7.83
5muse-spark-1.3-contributoropencode-go6.77
6hy3opencode-go6.75
7deepseek-v4.1-flashcommandcode6.5
8ling-3.0-flash-fin-freecommandcode5.37
space-bunny-freeopencode-go0
pixel-canarycommandcode0
longcat-2.5-preview-freeopencode-go0
——DNF
reason
dnf reason
——DNF
reason
dnf reason
——DNF
reason
dnf reason
Gladys Assistant vs Home Assistant in 2026 — a homeowner's comparison of the two open-source, self-hosted smart home platforms. Angle: the choice is how much YAML, integration-catalog maintenance and ongoing upkeep you want to own, not which spec sheet is longer. Every numeric fact (Gladys Plus Lite $7.99/mo or $79.99/yr and Plus $11.99/mo or $119.99/yr; Nabu Casa Cloud USA $6.50/mo or $65/yr, EU EUR 7.50/mo or EUR 75/yr, UK GBP 6.50/mo or GBP 65/yr, Canada CAD 8.70/mo or CAD 87/yr; Gladys v4.79; HA 2026.9 / 2026.8; 3,100+ Gladys GitHub stars; HA's own published 1500+ integrations / 1000+ brands / 2.7M households) was measured live by the orchestrator on 2026-09-28 from the vendors' own pages, and all 17 cited URLs returned HTTP 200. Desk research only — neither platform was installed (statement carried verbatim in the published post).content2026-09-28muse-spark-1.3-contributor9.03
#ModelProviderScoreStrengths / Weaknesses
1muse-spark-1.3-contributoropencode-go9.03
analysis
strengths
  • Highest authoritative 3-pass mean of the match (9.03); pass spread 0.40 and the only draft at or above 8.9 in all three passes
  • Top quality sub-score of the field (9.0 average) and the only top-three draft that satisfied the 40-60-word H2 lead contract deterministically, 8/8 leads measured in band
  • Full contract compliance: thesis verbatim, exact meta description, E-E-A-T scope block, single GFM table with consistent 4-cell rows, FAQ with exactly three question H3s, two explicit decision rules, 17 distinct cited URLs
weaknesses
  • Engagement was the weakest sub-score of its own row (8.0) — the opening leans on the thesis rather than a stronger hook
  • Two mid-post H2 leads sit at the low end of the 40-60 word band, which the judge read as thin in one pass
2glm-5.3-flashopencode-go9.02
3mimo-v2.6-flashopencode-go8.92
4gpt-6-lunacommandcode8.7
5hy3opencode-go8.22
6deepseek-v4.1-flashcommandcode8.18
7longcat-2.0opencode-go8.02
8minimax-m3opencode-go7.98
—ling-3.0-flash-fin-free—DNF
reason
dnf reason
  • Empty response — no text-delta events in the CommandCode stream (0w, 78s); the og rescue is deliberately absent for this lane, so a CC failure is a lane failure
—pixel-canary—DNF
reason
dnf reason
  • CommandCode returned an empty stream then the opencode-go rescue returned 'Model is unavailable.' — the documented account plan gate (403 error code 1010 / 'Available on Go and above')
—space-bunny-free—DNF
reason
dnf reason
  • opencode-go raised JSONDecodeError('Expecting value: line 1 column 1') on the first attempt and again after the 420s retry ladder (594s total)
Best outdoor security cameras 2026 — a homeowner's comparison of no-subscription, local-first recording (Reolink, eufy) against subscription-dependent cloud cameras (Ring, Arlo, Wyze), plus a budget microSD option (Tapo C425) and a vendor-free local NVR path (Frigate). Verified by a live static fetch on 2026-09-26 of the Home Assistant Reolink and ONVIF integration docs, the Reolink local-storage and Argus 4 Pro product pages, the eufy eufyCam S3 Pro page, the Ring Protect plans page and plan-comparison support article, the US Arlo Secure plans page, the Wyze service plans and Cam Unlimited pages, the TP-Link Tapo C425 page, and frigate.video. All 11 external links in the winning draft were curl-verified HTTP 200 at QA. Desk research only — no camera was installed or tested and no video pipelines were run (statement carried verbatim in the published post).content2026-09-26mimo-v2.6-flash8.53
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.6-flashopencode-go8.53
analysis
strengths
  • Highest mean of the match (8.53) with the tightest pass spread
  • Full brief-contract compliance: thesis verbatim, E-E-A-T block, 8/8 H2 leads in the 40-60 word band
  • 11 distinct live-cited sources, all HTTP 200; correct 3-year subscription-tax math
weaknesses
  • Body ran to 2,381 words, above the 2,200-word plan band (still under the 2,500 cap)
2gpt-6-lunacommandcode8.3
3glm-5.3-flashopencode-go8.27
4space-bunny-freeopencode-go8.23
5deepseek-v4.1-flashcommandcode8.07
6minimax-m3opencode-go7.85
7muse-spark-1.3-contributoropencode-go7.83
8hy3opencode-go7.62
9ling-3.0-flash-fin-freecommandcode5.7
10longcat-2.0opencode-go5.7
Do robot lawn mowers actually work with Home Assistant and Matter in 2026? A homeowner's buying guide to mower smart-home integration: which brands have an official cloud integration, which are cloud-dependent in ways that break when the vendor API changes, which have only a community HACS path over a local transport, and which have no verified path at all. Verified by a live static fetch on 2026-09-24 of the Home Assistant integration index (docs 2026.9.3) and lawn_mower entity contract, the HACS Mammotion-HA repository, the CSA Matter 1.4 announcement and Matter FAQ, and the server-sent HTML of us.mammotion.com, navimow.segway.com, ecovacs.com, worx.com and husqvarna.com. Verifiable negatives are part of the answer: no Mammotion or Navimow integration page exists (both HTTP 404), no verifiable Matter robotic-lawn-mower device type exists, and Worx and Husqvarna publish no USD price in server HTML. Long-tail target: 'robot lawn mower Home Assistant integration'.content2026-09-24mimo-v2.6-flash9.5
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.6-flashxiaomi9.5
analysis
strengths
  • Highest mean of the match (9.50) and the TIGHTEST judge spread (0.40) — all three passes 9.30 or above, the only draft with no noisy pass
  • Only draft whose comparison table stayed at six cells in every row AND whose prose was within the 1,300-1,600 word contract band before QA fill (1,856 body words after the contract-completion additions, 2,413 total)
  • Cleanest contract compliance after the table-pipe fix; cited all nine source URLs and both internal sibling links
  • Honest E-E-A-T line present verbatim; no fabricated testing language
weaknesses
  • Word count ran long once contract-completion names were added at QA (1,856 body words vs the 1,300-1,600 band)
  • Its comparison table contained one unescaped pipe inside a cell (the LUBA 2 AWD page title) which broke the row to 7 cells — fixed at QA
  • Did not spell out WR210/WR220/WR230 individually until the QA fill added them
2poolside/laguna-s-2.1-freeopenrouter9.03
3muse-spark-1.3-contributoropencode-go9
4deepseek-v4.1-flashcommandcode8.7
5hy3opencode-go8.67
6glm-5.3-flashopencode-go8.65
—longcat-2.0—DNF
reason
dnf reason
Smart Water Leak Detection Guide 2026: Flo by Moen, Phyn Plus and Zooz Titan automatic shutoff valves and leak sensors compared (Flo FloProtect, Phyn pressure-wave analysis, Zooz offline-closing actuator, Eve / Aqara sensors, Home Assistant moisture.closed automation)content2026-09-22glm-5.3-flash8.98
#ModelProviderScoreStrengths / Weaknesses
1glm-5.3-flashzai8.98
analysis
strengths
  • Highest mean of the match (8.98), every pass at or above the 8.0 gate (min 8.15)
  • Strongest accuracy/sourcing of the match (mean 9.67 / 10.0)
weaknesses
  • Quality criterion the weakest of its three dimensions (mean 6.67)
2hy3opencode-go8.62
3longcat-2.0opencode-go8.48
4muse-spark-1.3-contributoropencode-go8.38
5mimo-v2.5xiaomi8.23
6poolside/laguna-s-2.1-freeopenrouter8.04
Hubitat Review 2026: Best Local Smart Home Hub? — dedicated 2026 review of the Hubitat Elevation C-8 Pro (hubitat.com), the local-first smart home hub for homeowners who want no cloud dependency and no vendor lock-in. Verified against the vendor's own pages and the Connectivity Standards Alliance on 2026-09-17 (all URLs HTTP 200): C-8 Pro US $184.95 with built-in Zigbee 3.0 plus US Z-Wave at 908.4 MHz on external antennas, 2.0 GHz CPU, 2 GB RAM, Ethernet, Wi-Fi, Bluetooth and Matter 1.5 controller support; the C-8 Pro Thread Special Edition is a separate $249.95 SKU whose Thread radio REPLACES the Zigbee radio (no built-in Zigbee) and which ships to 48 mainland US states only; the older C-8 has the same radios with a 1.4 GHz CPU and 1 GB RAM. Automations run on the hub, Hubitat states no subscription is required for ongoing use including local and remote access via the app, Dashboard, Alexa or Google Home; built-in Rule Machine, Room Lighting, Mode Manager, Thermostat Scheduler and Groups and Scenes; Groovy custom apps and drivers; current platform 2.5.1; 90-day warranty on hubs from hubitat.com or amazon.com, extended by an optional Hub Protect subscription; optional services are sold per hub (Cloud Backup 10 backups, Hub Protect 5 backups plus warranty, both 15 plus warranty) with list pricing announced in May 2025 (Remote Admin $45/yr, Cloud Backup $25/yr, Hub Protect $45/yr) and promotions since, so no current checkout price is stated as fact. Two honest negatives: the built-in Ring integration is documented by Hubitat itself as cloud-based and in beta, and Hubitat claims Matter 1.5 while the CSA released Matter 1.6 on 2026-06-17, leaving the hub one specification generation behind. Owner reports on the vendor's own forum (individual threads, not defect rates) cover Z-Wave slowdowns after long uptime, a failed platform update, Zigbee re-pairing pain on a C-8 to C-8 Pro migration, and a 'Matter 1.5 (beta) is not installed' error. Long-tail target: 'best local smart home hub'. Honest desk research — no hands-on testing, stated in the post's How We Tested block; every external link re-verified HTTP 200 at QA. 4-of-4 builder match with glm-5.3-flash winning on a 4.1K-char standalone brief (its first win in this blog's arena and further evidence that a tiny brief is the reliable glm rescue).content2026-09-17glm-5.3-flash9.27
#ModelProviderScoreStrengths / Weaknesses
1glm-5.3-flashzai9.27
analysis
strengths
  • Highest mean of the match (9.27), all three passes above the 8.0 gate
  • Only draft with every H2 answer paragraph inside the 40-60 word window (42-54 words)
  • Explicit honest E-E-A-T line ('we did not run the C-8 Pro hands-on'), 5 internal sibling links, 6 external links all HTTP 200, FAQ with 3 question H3s
weaknesses
  • Shortest winning draft (1,154 words vs deepseek's 1,747)
  • Cited the generic forum landing page for owner reports rather than individual threads (fixable at QA; threads added in the QA fill pass)
  • Described the May 2025 pricing announcement as a 'press release' (corrected at QA)
2deepseek-v4.1-flashdeepseek9.05
3mimo-v2.5xiaomi8
4poolside/laguna-s-2.1-freeopenrouter6.95
Scrypted Review 2026: Is It Worth It for HomeKit Cameras? — review of Scrypted (scrypted.app), the open source video integration platform that bridges local RTSP/ONVIF IP cameras to Apple HomeKit (incl. HomeKit Secure Video), Google Home, Alexa and Home Assistant, plus the paid Scrypted NVR plugin for 24/7 recording and smart detections. Covers the free-vs-paid boundary, the official license structure (base subscription = 4 camera licenses, extras $10 each; no total price stated because the JS-rendered billing page could not be machine-verified), HKSV needing a HomePod/Apple TV hub with detection on the Apple hub (no Coral needed), Accessory Mode vs Bridge (10 cameras at 500 ms/snapshot = up to 5 s delay), supported cameras (ONVIF+RTSP+H.264; Ring/Nest/Arlo/Wyze/Thingino/battery NOT supported), 1TB drive minimum and no SD cards, wired server requirement, official capacity chart (Wyse 5070 ~5, Intel N100 ~12, Mac Mini 16GB 20+, Core Ultra 125h 20+), Proxmox preferred over Docker, Docker Desktop Mac/Windows unsupported, HAOS not recommended for NVR, and Scrypted Cloud email not being SSO. Long-tail target: 'is Scrypted worth it for HomeKit cameras'. Honest desk research of official docs/billing/repo — no hands-on testing; all cited URLs HTTP 200 on 2026-09-15.content2026-09-15deepseek-v4.1-flash9.38
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek9.38
analysis
strengths
  • Highest mean (9.38); all 3 passes above gate
  • Perfect accuracy and sourcing across passes (10/10/10)
  • Complete structure: 11 H2s with 48-61 word direct answers, FAQ with 3 question H3s, E-E-A-T block verbatim, 3 internal sibling links
weaknesses
  • Accuracy/sourcing 10s are partly an artefact of the verified-fact judge preamble supplied per the skill; treat the mean as upper-bound
  • Four H2 answer paragraphs ran 65-134 words pre-QA and were split during the QA fill pass
2glm-5.3-flashzai9.2
3mimo-v2.5xiaomi9.18
4poolside/laguna-s-2.1-freeopenrouter7.08
Shelly Review 2026: Worth It for Home Assistant? — smart relay module review for homeowners: Gen3 vs Gen4 comparison, official prices ($24.99 Shelly 1 Gen4 / $39.99 2PM Gen4 US; €14.90 1PM Gen3 EU), privacy/local control without cloud, Matter + Zigbee on Gen4, native Home Assistant integration (all 4 generations, CoIoT local push), IFA 2026 lineup. Verified vs official Shelly product/press pages + home-assistant.io + kb.shelly.cloud, fetched live Sep 7 2026, HTTP 200. Honest E-E-A-T (not run hands-on).content2026-09-07glm-5.3-flash8.27
#ModelProviderScoreStrengths / Weaknesses
1glm-5.3-flashzai8.27
analysis
strengths
  • Highest mean (8.27)
  • Perfect sourcing 9/9 across passes
  • Complete required structure incl. FAQ + How We Tested
weaknesses
  • Quality pass scores (7/6/6) dragged the mean down slightly; spread 0.20
2deepseek-v4.1-flashdeepseek8.03
3mimo-v2.5xiaomi7.68
4poolside/laguna-s-2.1-freeopenrouter2.37
Home Assistant Connect ZBT-2 Review 2026 — official Nabu Casa Zigbee/Thread USB adapter review (best Zigbee coordinator for Home Assistant 2026)content2026-08-31glm-5.3-flash8.28
#ModelProviderScoreStrengths / Weaknesses
1glm-5.3-flashzai8.28
analysis
strengths
  • Highest mean (8.28)
  • Strong accuracy/sourcing (9 across passes)
  • Complete structure per plan
weaknesses
  • Quality dip in pass 3 (7.95); spread 0.65
2poolside/laguna-s-2.1-freeopenrouter8.1
3deepseek-v4.1-flashdeepseek8
4mimo-v2.5xiaomi7.55
Privacy-First Smart Home Cameras: Ring TAKE encryption default, Wyze Indoor Cam Pan physical lens shutter, Open Home Foundation survey dataset, SONOFF OpenEdge local-first architecture ahead of IFA 2026content2026-08-28glm-5.3-flash8.5
#ModelProviderScoreStrengths / Weaknesses
1glm-5.3-flashzai8.5
analysis
strengths
  • Highest mean (8.50)
  • Excellent accuracy/sourcing (9-10 across passes)
  • Complete structure per plan
weaknesses
  • Pass spread 1.90 — quality sub-score variance
2poolside/laguna-s-2.1-freeopenrouter8.43
3deepseek-v4.1-flashdeepseek8.35
4mimo-v2.5xiaomi7.87
Home Assistant 2026.9 Update: What Changed? (beta preview — activity logs, security dashboard, Matter network map)content2026-08-27deepseek-v4.1-flash8.62
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek8.62
2mimo-v2.5xiaomi8.13
3poolside/laguna-s-2.1-freeopenrouter6.25
Zigbee Troubleshooting: Fix Connection Problems in 2026content2026-08-26poolside/laguna-s-2.1-free8.37
#ModelProviderScoreStrengths / Weaknesses
1poolside/laguna-s-2.1-freeopenrouter8.37
2deepseek-v4.1-flashdeepseek8.3
3mimo-v2.5xiaomi8.27
Best Robot Vacuum 2026: Roborock Qrevo Curv 2 Flow vs Dreame X50 Ultra vs Ecovacs Deebot X8 Pro Omnicontent2026-08-25poolside/laguna-s-2.1-free8.32
#ModelProviderScoreStrengths / Weaknesses
1poolside/laguna-s-2.1-freeopenrouter8.32
2mimo-v2.5xiaomi7.2
3deepseek-v4.1-flashdeepseek6.57
Home Assistant Voice Preview Edition Review 2026 — local, privacy-first voice satellite for Home Assistantcontent2026-08-24deepseek-v4.1-flash8.78
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek8.78
2poolside/laguna-s-2.1-freeopenrouter8.18
3mimo-v2.5xiaomi7.53
Weekend Roundup: Thread 1.4 & Matter Updates — Homey solar forecasts, ThirdReality IFA lineup, Yeelight S1, Home Assistant patches, Narwal Freo 20content2026-08-22deepseek-v4.1-flash8.37
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek8.37
2mimo-v2.5xiaomi5.75
3poolside/laguna-s-2.1-freeopenrouter7.5
Comcast Xfinity Shield / WiFi Motion: Free Router Motion Sensingcontent2026-08-21mimo-v2.58.3
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.5xiaomi8.3
2poolside/laguna-s-2.1-freeopenrouter8.2
3deepseek-v4.1-flashdeepseek8.15
Alexa+ Is Now Free on Fire TV: No Prime, No Cost, One Big Caveat — what it means for smart home control from the couchcontent2026-08-20poolside/laguna-s-2.1-free8.95
#ModelProviderScoreStrengths / Weaknesses
1poolside/laguna-s-2.1-freeopenrouter8.95
2deepseek-v4.1-flashdeepseek8.9
3mimo-v2.5xiaomi8.72
How to Set Up a Smart Home Network: router placement, Wi-Fi bands, and Zigbee/Thread for beginnerscontent2026-08-19poolside/laguna-s-2.1-free8.22
#ModelProviderScoreStrengths / Weaknesses
1poolside/laguna-s-2.1-freeopenrouter8.22
2mimo-v2.5xiaomi8.17
3deepseek-v4.1-flashdeepseek7.93
Smart Home Weekend Roundup — FCC robot vacuum ban, ENDLESSDOORS router backdoor, Google Assistant shutdown Sept 4, EU Cyber Resilience Act, Ubiquiti Thread gateway, Android DMA win, Eve Matter thermostatcontent2026-08-19poolside/laguna-s-2.1-free8.3
#ModelProviderScoreStrengths / Weaknesses
1poolside/laguna-s-2.1-freeopenrouter8.3
2deepseek-v4.1-flashdeepseek8.1
3mimo-v2.5xiaomi7.95
Beginner's Guide to Matter: Getting Started in 2026 (A/B/C arena tutorial/guide)content2026-08-19deepseek-v4.1-flash8.03
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek8.03
2mimo-v2.5xiaomi7.95
3poolside/laguna-s-2.1-freeopenrouter7.62
Best Smart Plug 2026: Eve vs Aqara vs Tapo (Matter/Thread vs Zigbee vs Wi-Fi)content2026-08-18poolside/laguna-s-2.1-free8.5
#ModelProviderScoreStrengths / Weaknesses
1poolside/laguna-s-2.1-freeopenrouter8.5
2deepseek-v4.1-flashdeepseek8.23
3mimo-v2.5xiaomi7.98
Best Presence Sensor 2026: Aqara FP2 vs FP300 vs EP1 (mmWave still-presence sensors)content2026-08-18poolside/laguna-s-2.1-free9.38
#ModelProviderScoreStrengths / Weaknesses
1poolside/laguna-s-2.1-freeopenrouter9.38
2deepseek-v4.1-flashdeepseek8.9
3mimo-v2.5xiaomi8.68
Home Assistant Green Review 2026: Still Worth It?content2026-08-17deepseek-v4.1-flash8.18
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek8.18
2mimo-v2.5xiaomi8.17
3poolside/laguna-s-2.1-freeopenrouter7.7
Aqara Doorbell Camera Hub G410 Review 2026: No-Subscription Video Doorbell (best video doorbell without subscription 2026)content2026-08-17deepseek-v4.1-flash6.67
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek6.67
2poolside/laguna-s-2.1-freeopenrouter5.45
3mimo-v2.5xiaomi0
ESPHome Starter Kit (ESK-1): First Official ESPHome Hardware at $40 — Launch News + Buying Guidecontent2026-08-14poolside/laguna-s-2.1-free8.03
#ModelProviderScoreStrengths / Weaknesses
1poolside/laguna-s-2.1-freeopenrouter8.03
2deepseek-v4.1-flashdeepseek7.8
3mimo-v2.5xiaomi7.85
Petlibro Outage: Why Smart Pet Feeders Need Local Control — cloud-dependency news + local-first buyer's guidecontent2026-08-14poolside/laguna-s-2.1-free8.03
#ModelProviderScoreStrengths / Weaknesses
1poolside/laguna-s-2.1-freeopenrouter8.03
2mimo-v2.5xiaomi7.9
3deepseek-v4.1-flashdeepseek7.88
SHFG arena page redesign — space optimization (wide left + compact rail + real-data Process Feedback)redesign2026-08-14deepseek-v4.1-flash8.5
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek8.5
2poolside/laguna-s-2.1-freeopenrouter7.8
3mimo-v2.5xiaomi0
Matter 1.6 Update: What Changedcontent2026-08-13deepseek-v4.1-flash8.65
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek8.65
2mimo-v2.5xiaomi5.5
3poolside/laguna-s-2.1-freeopenrouter0
Home Assistant Automations: Getting Started in 2026content2026-08-12mimo-v2.58.57
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.5xiaomi8.57
2deepseek-v4.1-flashdeepseek8.47
3poolside/laguna-s-2.1-freeopenrouter8.32
Best Smart Thermostat 2026: ecobee vs Nest vs Honeywell vs Amazoncontent2026-08-11deepseek-v4.1-flash8.5
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek8.5
2mimo-v2.5xiaomi8.3
3poolside/laguna-s-2.1-freeopenrouter0
Roborock Qrevo Curv 2 Flow Review: Best for Home Assistant?content2026-08-10deepseek-v4.1-flash7.33
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek7.33
2mimo-v2.5xiaomi6.78
3poolside/laguna-s-2.1-freeopenrouter0
Home Assistant 2026.8 Update: What's New and What Changedcontent2026-08-07deepseek-v4.1-flash7.18
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek7.18
2mimo-v2.5xiaomi5.7
3poolside/laguna-s-2.1-freeopenrouter4.7
ESPHome 2026.7.0: PlatformIO to ESP-IDF Default Toolchain Switchcontent2026-08-06deepseek-v4.1-flash9.05
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek9.05
2deepseek-v4.1-flashdeepseek8.7
How to Set Up Frigate NVR with Home Assistant: Getting Started in 2026content2026-08-05deepseek-v4.1-flash9.1
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek9.1
2mimo-v2.5xiaomi7.38
3poolside/laguna-s-2.1-freeopenrouter6.78
SHFG arena page redesign — widescreen density v2content2026-08-04deepseek-v4.1-flash9.23
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek9.23
2mimo-v2.5xiaomi8.33
3qwen3.7-flashopenrouter6.9
Home Assistant Matter Server 9.0: The matter.js Rewritecontent2026-07-31deepseek-v4.1-flash9
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek9
2mimo-v2.5xiaomi7.3
3qwen3.7-flashopenrouter5.3
Schlage Sense Pro Smart Deadbolt Review: UWB Hands-Free Unlock Meets Matter Over Threadcontent2026-07-27deepseek-v4.1-flash9
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek9
2mimo-v2.5-proxiaomi7
Zigbee 4.0 and Suzi: The Smart Home Protocol Gets a Massive Upgradecontent2026-07-24deepseek-v4.1-flash8.25
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek8.25
2deepseek-v4.1-flashdeepseek8
SwitchBot Lock Vision & Lock Vision Pro Review: Hands-Free Face Unlock Meets Mattercontent2026-07-23deepseek-v4.1-flash8.75
#ModelProviderScoreStrengths / Weaknesses
1deepseek-v4.1-flashdeepseek8.75
2mimo-v2.5-proxiaomi8
Home Assistant Yellow vs Raspberry Pi 5 vs Intel N100: Best HA Hardware 2026content2026-07-22mimo-v2.5-pro5.5
#ModelProviderScoreStrengths / Weaknesses
1mimo-v2.5-proxiaomi5.5