Skip to content
Undominated.ai
/
LeaderboardRangesFrontierCheckTools

429 models · 52 providers · 6 on the frontier · as of Sep 19, 2026

Check a model

Decide

Quality first, then price.

LeaderboardRangesFrontierCheckCompareImagesVideoEmbeddings

Pay

Who sells it, and the price.

PlansCheapest atProvidersSpreadsAuditFreeOpenRouter or directCheaper alternatives

Run

Tools, weights, hardware.

ToolsSelf-hostOpen weightsFamiliesBatchCacheDeprecationsFrontier embed

Measure

Cliffs, knees, effort.

CliffsSignificanceQuality kneeReasoning effortBlind rank testTrapsDisagreementBest

Trust

Method, log, corrections.

NowMethodologyCorrectionsIndependenceReport a wrong pricePress kitThe wireCite this dataset

Open

The index, the JSON, the tracker.

All modelsCheck a repoAPIllms.txtThis page as MarkdownPublic GitHubFeedllms-full.txt
  1. Leaderboard /
  2. Providers /
  3. Meta

Providers

Meta

15 models from Meta in this catalogue, as of 2026-09-19. Unrated is not a score of zero.

  • Llama 3.1 70B Instruct
  • Llama 3.1 8B Instruct
  • Llama 3.2 1B Instruct
  • Llama 3.2 3B Instruct
  • Llama 3.3 70B Instruct
  • Llama 4 Maverick unrated
  • Llama 4 Scout unrated
  • Llama Guard 4 12B unrated
  • Muse Glimmer 30B unrated
  • Muse Glimmer 30B (batch) (batch) unrated
  • Muse Spark 1.1
  • Muse Spark 1.2 unrated
  • Muse Spark 1.2 Contributor unrated
  • Muse Spark 1.3
  • Muse Spark 1.3 Contributor unrated
Undominated.ai

Best first. Then price.

Independent ranking. Quality first, then what it costs.

Check a model Report a wrong price
models
429
providers
52
on the frontier
6
ranked with a price
137
as of
Sep 19, 2026

Decide

Quality first, then price.

Leaderboard Ranges Frontier Check Compare Images Video Embeddings

Pay

Who sells it, and the price.

Plans Cheapest at Providers Spreads Audit Free OpenRouter or direct Cheaper alternatives

Run

Tools, weights, hardware.

Tools Self-host Open weights Families Batch Cache Deprecations Frontier embed

Measure

Cliffs, knees, effort.

Cliffs Significance Quality knee Reasoning effort Blind rank test Traps Disagreement Best

Trust

Method, log, corrections.

Now Methodology Corrections Independence Report a wrong price Press kit The wire Cite this dataset

Open

The index, the JSON, the tracker.

All models Check a repo API llms.txt This page as Markdown Public GitHub Feed llms-full.txt

Published prices and independent scores are source-linked. Derived ranks, workload prices, break-evens, and hardware-fit figures name their assumptions. Artificial Analysis figures are not published here: we have no redistribution licence, so they inform what we check and never what you read.

Built and maintained by Len van der Hof