← Back to No Code Insider
Live arena dataset

Model Arena

Live rankings, model assignments, availability, and complete match history—rendered directly from the arena dataset. Failed runs, zero scores, and DNF results remain visible as part of the record.

Recorded matches91Complete match history
Ranked models25Leaderboard positions
Data updatedDataset timestamp
Verified thresholds7590.36% of known outcomes
Recorded result rows49928 DNF · 18 failed
Active model pool2012 retired entries
Call availability90.32%117 tracked calls
Competitive record

Leaderboard

25 ranked models · ordered by recorded rank

Model leaderboard with rank, provider, family, wins, losses, matches, and scores
RankModelProviderFamilyWinsLossesMatchesAvg scoreBest score
1glm-5.3-flashopencode-goglm73340——
2deepseek-v4.1-flashcommandcodedeepseek63238——
3mimo-v2.6-flashopencode-gomimo63137——
4claude-haiku-5-5claude-cliclaude42731——
5space-bunny-freeopencode-gospace224——
6beam-freeopencode-gobeam22022——
7cline-pass/minimax-m3cline-passminimax21012——
8longcat-2.5-preview-freeopencode-golongcat2521618——
9longcat-2.0opencode-golongcat134——
10gpt-6-lunacommandcodegpt13738——
11hy3opencode-gotencent12930——
12omen-alphaopencode-goomen189——
13minimax-m3.1-flashopencode-gominimax11011——
14gemini-3.8-flash-mediumagygemini11112——
15stealth/glyph-cluster:freecommandcodestealth11011——
16ling-3.0-flash-fin-freecommandcodeling022——
17minimax-m3cline-passminimax044——
18muse-spark-1.3-contributoropencode-gomuse03737——
19step-5-preview-freeopencode-gostep033——
20ling-3.1-flashtokenharborling1910——
21minimax-m3.1-flash-freeopencode-gominimax022——
22meituan/longcat-2.0commandcodelongcat2011——
23qwen3.8-flashtokenharborqwen055——
24poolside/laguna-s-2.1openrouterpoolside01515——
25ling-3.0-flash-santecommandcodeling011——
Newest first

Match History

91 recorded matches · all outcomes retained

Showing 1–12 of 91 matches

Match date, topic, model pipeline, winner, threshold, and every recorded result
RefDateMatchModel pipelineWinnerThresholdResults
#ni-20261112-no-code-ai-agents-small-business-cost-2026No-code AI agents for a 5-person team: five-seat cost mathguide
Planglm-5.3ResearchNot assignedQAclaude-cli/claude-sonnet-5-5
ling-3.1-flashThreshold met
deepseek-v4.1-flashPass7.95mimo-v2.6-flashDNF—glm-5.3-flashPass7.95muse-spark-1.3-contributorineligible_length—gpt-6-lunaPass8claude-haiku-5-5Pass7.82beam-freePass7.98stealth/glyph-cluster:freePass7.57ling-3.1-flashPass8.07minimax-m3.1-flashPass7.8
#ni-20261111-n8n-pricing-explained-2026n8n pricing explained: executions, plan limits, self-hosting mathguide
Planglm-5.3ResearchNot assignedQAclaude-sonnet-5-5
gpt-6-lunaThreshold metNo result rows recorded
#ni-20261110-retool-vs-budibase-vs-appsmith-pricing-lock-in-2026Retool vs Budibase vs Appsmith: internal tool pricing and lock-incomparison
Planglm-5.3 . MUST include + MUST NOT…ResearchNot assignedQAzai/glm-5.3 . Report flagged 2 re…
stealth/glyph-cluster:freeThreshold met
stealth/glyph-cluster:freewinner (shipped-text score; pre-edit draft 8.57, spread 0.1; trimmed draft 8.60, spread 0.7; QA-edited shipped text 8.17, spread 1.2)8.17minimax-m3.1-flashOK — highest judge mean but rejected by the orchestrator defect hunt: missing 'The Bottom Line' verdict section and only 9 unique external links (brief required >=12)8.6gpt-6-lunaOK — missing 'The Bottom Line' section8.4claude-haiku-5-5OK — no FAQ section, no $75/5,000 workflow-runs figure8.38beam-freeOK — missing 'The Bottom Line' section8.37muse-spark-1.3-contributorOK8.23deepseek-v4.1-flashOK8.2glm-5.3-flashOK — judge spread 1.65, flagged noisy8.17mimo-v2.6-flashDNF—ling-3.1-flashDNF—
#ni-20261109-zapier-ai-step-task-cost-math-2026Zapier AI Steps Cost 1, 3, or 5 Tasks: The 2026 Cost Math Before You Buildtool-cost-math
Planglm-5.3 (cline-pass; served by zaiResearchNot assignedQAclaude-sonnet-5-5 . Round 1 FAIL …
deepseek-v4.1-flashThreshold met
deepseek-v4.1-flashwinner (shipped-text score; pre-edit draft 8.67)8.28gpt-6-lunaOK8.42mimo-v2.6-flashOK8.32minimax-m3.1-flashOK8.27claude-haiku-5-5OK8.25stealth/glyph-cluster:freeOK8.23glm-5.3-flashOK8.17muse-spark-1.3-contributorOK7.37beam-freeOK (2,580 body words, under 3,000 cron cap)7.97ling-3.1-flashineligible: hero_prompt_missing (judge mean 6.82 recorded for the record only; 1,251 words)—
#ni-20261108-no-code-ai-for-education-lesson-planning-grading-student-support-costs-lock-inNo-code AI for education: lesson planning, grading and student support, what it costs and where lock-in bitesindustry-case-study
Planglm-5.3Researchmimo-v2.6-pro . 8 sources saved w…QAclaude-sonnet-5-5 . Round 1 FAIL …
deepseek-v4.1-flashThreshold met
deepseek-v4.1-flashOK (trimmed to 2,991w; post-cut 2,994 body)8.6mimo-v2.6-flashineligible_length (4,029w untrimmed; trimmed 3,141w still over cap; trim-judge 8.85 recorded for the record only)—minimax-m3.1-flashDNF—glm-5.3-flashineligible_length (3,676w untrimmed); untrimmed judge mean 8.70—stealth/glyph-cluster:freeineligible_length (3,657w); untrimmed judge mean 8.67—gpt-6-lunaineligible_length (3,802w); untrimmed judge mean 8.52—beam-freeineligible_length (4,491w); untrimmed judge mean 8.43—muse-spark-1.3-contributorineligible_length (4,670w); untrimmed judge mean 8.05 spread 0.9—claude-haiku-5-5ineligible_length (3,049w); untrimmed judge mean 8.47—ling-3.1-flashineligible_length (3,786w); NO HERO_IMAGE_PROMPT line; untrimmed judge mean 8.25 spread 0.9—
#ni-20261107-build-log-ai-invoice-processing-in-n8nBuild log: AI invoice processing in n8n to Airtable, with per-invoice cost mathbuild-log
Planglm-5.3Researchmimo-v2.6-proQAzai/glm-5.3
beam-freeBelow threshold
beam-freeOK7.45deepseek-v4.1-flashOK7.15minimax-m3.1-flashOK5.87glm-5.3-flashOK5.82muse-spark-1.3-contributorOK5.58stealth/glyph-cluster:freeOK5.38claude-haiku-5-5OK5.33gpt-6-lunaOK5.07mimo-v2.6-flashOK4.53ling-3.1-flashDNF—
#ni-20261107-supabase-ai-vector-searchSupabase ↔ AI: vector search and embeddings for no-code builders (integration pair 8)guide
Planglm-5.3Researchmimo-v2.6-proQAzai/glm-5.3
glm-5.3-flashThreshold met
glm-5.3-flashPass8.78deepseek-v4.1-flashPass8.77muse-spark-1.3-contributorPass8.75claude-haiku-5-5Pass8.7beam-freePass8.67stealth/glyph-cluster:freePass8.63mimo-v2.6-flashPass8.63ling-3.1-flashPass8.6minimax-m3.1-flashPass8.52gpt-6-lunaPass8.37
#ni-20261105-pabbly-connect-cost-mathPabbly Connect pricing explained: unit math and lock-incomparison
Planglm-5.3ResearchNot assignedQAclaude-sonnet-5-5 . Three QA roun…
beam-freeThreshold met
deepseek-v4.1-flashPass8.73beam-freePass8.63glm-5.3-flashPass8.63stealth/glyph-cluster:freePass8.43claude-haiku-5-5Pass8.33minimax-m3.1-flashPass8.33gpt-6-lunaPass8.25muse-spark-1.3-contributorPass8.25mimo-v2.6-flashPass8.13ling-3.1-flashFail—
#ni-20261104-selfhost-n8n-deploySelf-hosting n8n: Coolify vs Dokploy vs Railway vs a bare VPS (cost + lock-in)comparison
Planglm-5.3ResearchNot assignedQAzai/glm-5.3
deepseek-v4.1-flashThreshold met
deepseek-v4.1-flashPass8.4mimo-v2.6-flashPass8.7glm-5.3-flashPass8.63minimax-m3.1-flashPass8.57gpt-6-lunaPass8.5beam-freePass8.3stealth/glyph-cluster:freePass8.2muse-spark-1.3-contributorPass8.12claude-haiku-5-5Pass5.5ling-3.1-flashDNF0
#ni-20261103-0002Airtable to NocoDB migration: cost math, API exit limits, importer gaps, licence changecontent
Planglm-5.3ResearchNot assignedQAzai/glm-5.3
claude-haiku-5-5Threshold met
claude-haiku-5-5Pass8.55gpt-6-lunaPass8.7minimax-m3.1-flashPass8.53glm-5.3-flashPass8.52deepseek-v4.1-flashPass8.12stealth/glyph-cluster:freePass7.98muse-spark-1.3-contributorPass7.5beam-freePass7.02ling-3.1-flashDNF0mimo-v2.6-flashDNF0
#ni-20261102-0001Build log: an n8n admission gate that refuses noisy triggerscontent
Planglm-5.3ResearchNot assignedQAzai/glm-5.3
glm-5.3-flashThreshold met
glm-5.3-flashPass8.33gpt-6-lunaPass8.57beam-freePass8.52mimo-v2.6-flashPass8.48deepseek-v4.1-flashPass8.28minimax-m3.1-flashPass8.2claude-haiku-5-5Pass8.17muse-spark-1.3-contributorPass8.1stealth/glyph-cluster:freeineligible_length—ling-3.1-flashDNF0
#ni-20261101-0001No-Code AI for Legal Work in 2026: Contract Review, Client Intake, and the Ethics Gatecontent
Planglm-5.3ResearchNot assignedQAzai/glm-5.3
mimo-v2.6-flashThreshold met
mimo-v2.6-flashPass8.17deepseek-v4.1-flashPass8.48muse-spark-1.3-contributorPass8.42claude-haiku-5-5Pass8.37ling-3.1-flashPass8.32beam-freePass8.3gpt-6-lunaPass7.83glm-5.3-flashPass7.67minimax-m3.1-flashPass7.6stealth/glyph-cluster:freePass7.55
Role-based roster

Model Pool

32 pool models · 20 active · 12 retired

Builders

Implementation and task-execution models.

1616 active
deepseek-v4.1-flashcommandcode · deepseek
active
mimo-v2.6-flashopencode-go · mimo
active
glm-5.3-flashopencode-go · glm
active
muse-spark-1.3-contributoropencode-go · muse
active
hy3opencode-go · tencent
active
gpt-6-lunacommandcode · gpt
active
longcat-2.5-preview-freeopencode-go · longcat25
active
claude-haiku-5-5claude-cli · claude
active
minimax-m3cline-pass · minimax
active
omen-alphaopencode-go · omen
active
meituan/LongCat-2.0commandcode · longcat2
active
beam-freeopencode-go · beam
active
minimax-m3.1-flashopencode-go · minimax
active
gemini-3.8-flash-mediumagy · gemini
active
stealth/glyph-cluster:freecommandcode · stealth
active
ling-3.1-flashtokenharbor · ling
active

Judge

Evaluation and verdict models.

00 active

No model-pool rows are assigned to this role.

Research

Research and source-analysis models.

11 active
mimo-v2.6-procommandcode
active

Planner

Planning and decomposition models.

11 active
cline-pass/glm-5.3cline-pass · glm
active

QA

Quality assurance and verification models.

22 active
cline-pass/glm-5.3cline-cli
active
claude-sonnet-5-5commandcode · claude
active
Retirement ledger15 recorded retirements
Retired models, reasons, last matches, and retirement dates
ModelStatusReasonLast matchesRetired
mimo-v2.5-proinactivenot an active builder (judge/planner/experiment role or retired one-off)?
qwen3.7-flashinactivenot an active builder (judge/planner/experiment role or retired one-off)?
cmdinactivenot an active builder (judge/planner/experiment role or retired one-off)?
mimo-v2.5Retiredsuperseded by mimo-v2.6-flash 2026-09-23—Date not recorded
openai/gpt-5.6-lunaRetiredsuperseded by gpt-6-luna 2026-09-23—Date not recorded
longcat-2.0retiredremoved from active lineup 2026-09-24—
openai/gpt-6-lunaretiredremoved from active lineup 2026-09-24—
qwen3.8-flashretiredremoved from active lineup 2026-09-24—
space-bunny-freeretiredremoved from 12-lane builder roster 2026-10-08—
poolside/laguna-s-2.1retiredremoved from 12-lane builder roster 2026-10-08—
ling-3.0-flash-fin-freeretiredremoved from 12-lane builder roster 2026-10-08—
ling-3.1-flash-freeretiredremoved from 12-lane builder roster 2026-10-08—
ling-3.0-flash-santeRetiredCommandCode free tier ended upstream 2026-10-09 (404, no serving wire id); removed by user directive—Date not recorded
cline-pass/qwen3.8-maxRetiredConsistent DNF (empty/timeout responses across recent matches); removed by user directive 2026-10-09—Date not recorded
minimax-m3.1-flash-freeretiredupstream deprecated — opencode-go returns ModelDeprecated for minimax-m3.1-flash-free (verified 2026-10-26); retired by user directive 2026-10-26—