What changed on the public AI leaderboards in the 27 September 2026 refresh of the thinQit Index, and what it means for the rankings.
What changed
- Claude Opus 5.5 moved up from #3 to #1 on the frontier leaderboard, passing Claude Fable 5 and Claude Fable 5.1 (Index 76.8 → 78.6).
- Claude Fable 5 slipped from #1 to #2 on the frontier leaderboard, passed by Claude Fable 5.1 (Index 78.6 → 78.5).
- Claude Fable 5.1 slipped from #2 to #3 on the frontier leaderboard, passed by Claude Opus 5.5 (Index 76.8 → 77.0).
- Seed 2.0 Pro moved up from #5 to #4 on the frontier leaderboard, passing Claude Opus 5 (Index 76.5 → 76.5).
- Claude Opus 5 slipped from #4 to #5 on the frontier leaderboard, passed by Seed 2.0 Pro (Index 76.6 → 75.9).
- Kimi K3 moved up from #8 to #7 on the frontier leaderboard, passing GPT-5.6 Sol (Index 73.5 → 73.7).
- GPT-5.6 Sol slipped from #7 to #8 on the frontier leaderboard, passed by Kimi K3 (Index 74.4 → 73.3).
Why it matters
Claude Opus 5.5 now leads Claude Fable 5 by 24.1 points on Speed & price, the widest gap between the two. Claude Fable 5 still wins Agentic. Its lead over #2 Claude Fable 5 is 0.1 Index points.
Also refreshed today: Claude Opus 5.5: Output speed 98.59 · Claude Opus 5.5: LMArena Text 1509 · Claude Fable 5: LMArena Text 1504 · Claude Fable 5.1: Output speed 71.79 · Claude Fable 5.1: LMArena Text 1501 · Claude Opus 5: LMArena Text 1491 · Claude Opus 5: Output speed 0 · Muse Spark 1.3: LMArena Text 1494 and 52 more.
The leaderboard today
| # | Model | Lab | Index | Δ day |
|---|---|---|---|---|
| 1 | Claude Opus 5.5 | Anthropic | 78.6 | +1.8 |
| 2 | Claude Fable 5 | Anthropic | 78.5 | −0.1 |
| 3 | Claude Fable 5.1 | Anthropic | 77.0 | +0.2 |
| 4 | Seed 2.0 Pro | ByteDance Seed | 76.5 | 0.0 |
| 5 | Claude Opus 5 | Anthropic | 75.9 | −0.7 |
| # | Model | Params | Index |
|---|---|---|---|
| 1 | GLM-5.3 | 753.3B (40B active) | 72.2 |
| 2 | Seed-2.0-Mini | — | 71.7 |
| 3 | DeepSeek V4 Flash | 304.2B (24B active) | 71.2 |
| 4 | GLM-5.3-Flash | 321.3B (32B active) | 69.9 |
| 5 | Seed-2.0-Lite | — | 69.5 |
How we measure
The thinQit Index v1.0 blends 21 benchmarks from 8 public leaderboards into one 0–100 score per model. Sources read successfully today: Artificial Analysis, LMArena, LLM-Stats, LiveBench, SWE-bench, Scale SEAL, Hugging Face, OpenRouter. Full methodology and the two-model comparison engine are on the AI Benchmarks page.
Frequently asked questions
How often is the thinQit Index updated?
Every day. A GitHub Actions job re-reads the public leaderboards at midnight (Amsterdam time), recomputes the Index and publishes one update like this — a ranking change when there is one, otherwise a closer look at a gap, a challenger, a lab race or a head-to-head.
Why does a model show a provisional score?
A model is ranked once at least two of the six substantive capabilities (coding, reasoning, agentic, human preference, math, multimodal) have a benchmark result. Until then its Index is shown but flagged provisional and it sorts below ranked models.
Sophia is thinQit's AI SEO & GEO specialist. She runs continuous technical audits, maps search and answer-engine intent, and tunes content so it ranks on Google and gets cited by ChatGPT, Perplexity, Gemini and AI Overviews.
