Coding Division
Ranked by SWE-bench Verified: real GitHub issues, resolved end-to-end.
Table →The League
League standings & AI power rankings.
Models as teams: the AI Pro League table, ranked by our published composite of public benchmarks. This is the Season Zero form table — benchmarks today, benchmark + crowd once voted fixtures launch.
What changed
- Oct 15, 2026: Muse Wonder (Meta) added to the database — released 2026-10-12 at $4/$20 per 1M tokens, LiveBench 78.1.
- Oct 15, 2026: Mistral Large 4 LiveBench score corrected to 47.7 per LiveBench 2.0 (livebench.ai), replacing the 71.8 carried from a stale mirror leaderboard; League Score recomputed.
- Oct 14, 2026: Model database launched: 42 models tracked with sourced pricing and benchmarks, full League standings published.
- Oct 14, 2026: AI Labs section launched: 18 lab profiles, the Labs Leaderboard ranked by Club Score, Lab vs Lab comparisons and a per-lab release tracker.
- Oct 14, 2026: AI Movies section launched: 32 films about AI with original reviews, Entertainment and AI Realism scores, and verified official trailers.
40Ranked models
Claude Fable 5.1League leader
5Benchmark sources
60/40Benchmark/crowd split, from voted fixtures
Premier table — top 20
| # | Model | Maker | League Score | Context | $/1M in/out | Weights |
|---|---|---|---|---|---|---|
| 1 | Claude Fable 5.1 | Anthropic | 97 | 1M | $10.00/$50.00 | Closed |
| 2 | GPT-6 Astra | OpenAI | 97 | 1.1M | $10.00/$50.00 | Closed |
| 3 | Claude Opus 5.5 | Anthropic | 96 | 1M | $4.00/$20.00 | Closed |
| 4 | GPT-6.1 Sol | OpenAI | 94 | 1.1M | $2.00/$10.00 | Closed |
| 5 | Claude Opus 5 | Anthropic | 92 | 1M | $5.00/$25.00 | Closed |
| 6 | Gemini 4 Argon | Google DeepMind | 91 | 1M | $2.00/$10.00 | Closed |
| 7 | Claude Sonnet 5.5 | Anthropic | 87 | 1M | $2.00/$10.00 | Closed |
| 8 | GPT-5.6 Sol | OpenAI | 82 | 1.1M | $4.00/$20.00 | Closed |
| 9 | Muse Wonder | Meta | 76 | 1M | $4.00/$20.00 | Closed |
| 10 | DeepSeek V4.1 Flash | DeepSeek | 75 | 1M | $0.30/$1.20 | Open |
| 11 | Claude Sonnet 5 | Anthropic | 74 | 1M | $2.00/$10.00 | Closed |
| 12 | GPT-5.6 Terra | OpenAI | 74 | 1.1M | $2.00/$12.00 | Closed |
| 13 | Grok 4.3 | xAI (SpaceXAI) | 74 | 1M | $1.25/$2.50 | Closed |
| 14 | Gemini 3.5 Flash-Lite | Google DeepMind | 70 | 1M | $0.30/$2.50 | Closed |
| 15 | Kimi K2.6 | Moonshot AI | 70 | 262.1K | $0.95/$4.00 | Open |
| 16 | Muse Spark 1.3 | Meta | 69 | 1M | $1.25/$4.25 | Closed |
| 17 | Claude Haiku 5.5 | Anthropic | 68 | 1M | $0.10/$0.50 | Closed |
| 18 | Grok 4.7 | xAI (SpaceXAI) | 67 | 500K | $2.00/$6.00 | Closed |
| 19 | DeepSeek V4 Pro (0813) | DeepSeek | 63 | 1M | $1.32/$3.96 | Open |
| 20 | Gemini 3.8 Flash | Google DeepMind | 63 | 1M | $0.75/$3.75 | Closed |
Full model database How the score works
Divisions
Reasoning Division
Ranked by GPQA Diamond: graduate-level science that resists lookup.
Table →Value Division
League Score per blended dollar — capability per dollar, the table no one else publishes.
Table →Open-Weight Division
Open-weight models only, ranked by League Score.
Table →This week’s fixtures
Title clash — 1st vs 2nd
Claude Fable 5.1 vs GPT-6 Astra
Fixture card →Form clash — 3rd vs 4th
Claude Opus 5.5 vs GPT-6.1 Sol
Fixture card →Value champion vs League leader
Claude Haiku 5.5 vs Claude Fable 5.1
Fixture card →Open-weight championship
DeepSeek V4.1 Flash vs Kimi K2.6
Fixture card →The clubs
Models are the teams; the labs are the clubs that field them. Club Score = 0.6 × best model + 0.4 × top-three average.
#1 Anthropic
Club Score 95.8 · best: Claude Fable 5.1 · 4 models in the squad.
Club profile →#2 OpenAI
Club Score 94.4 · best: GPT-6 Astra · 6 models in the squad.
Club profile →#3 Google DeepMind
Club Score 84.2 · best: Gemini 4 Argon · 5 models in the squad.
Club profile →#4 DeepSeek
Club Score 73.0 · best: DeepSeek V4.1 Flash · 2 models in the squad.
Club profile →#5 xAI (SpaceXAI)
Club Score 72.3 · best: Grok 4.3 · 2 models in the squad.
Club profile →Full Labs Leaderboard Releases by lab Lab vs Lab OpenAI vs Anthropic OpenAI vs DeepMind Anthropic vs DeepMind