LLM Leaderboard
Ranked by Coding Index · source AA
AA composite — LiveCodeBench, SciCode, etc.
#1 by Coding Index
Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback)
81.6
Anthropic · in $10.0 / out $50.0 per 1M · 70 tok/s
| # | Model | Coding |
|---|---|---|
| 1 | Claude Fable 5.1 (Adaptive Reasoning, Max Effort, Default Fallback) | 81.6 |
| 2 | Claude Fable 5.1 (Adaptive Reasoning, Xhigh Effort, Default Fallback) | 80.7 |
| 3 | Claude Fable 5.1 (Adaptive Reasoning, High Effort, Default Fallback) | 79.1 |
| 4 | GPT-5.6 Sol (xhigh) | 78.3 |
| 5 | Claude Opus 5 (Adaptive Reasoning, Max Effort) | 78.0 |
| 6 | GPT-5.6 Sol (max) | 77.4 |
| 7 | GPT-5.6 Sol (high) | 77.2 |
| 8 | GPT-6 Astra (high) | 77.1 |
| 9 | Claude Fable 5.1 (Adaptive Reasoning, Medium Effort, Default Fallback) | 77.1 |
| 10 | Claude Opus 5 (Adaptive Reasoning, Xhigh Effort) | 77.0 |
| 11 | GPT-6 Astra (max) | 76.9 |
| 12 | Grok 4.6 (high) | 76.8 |
| 13 | GPT-6 Astra (medium) | 76.7 |
| 14 | GPT-5.6 Terra (max) | 76.7 |
| 15 | Claude Fable 5 (Adaptive Reasoning, Max Effort, Opus 4.8 Fallback) | 76.5 |
| 16 | Claude Opus 5 (Adaptive Reasoning, High Effort) | 76.5 |
| 17 | Muse Spark 1.3 (xhigh) | 76.5 |
| 18 | Muse Spark 1.3 (max) | 76.3 |
| 19 | Gemini 3.8 Flash (high) | 76.3 |
| 20 | GPT-5.6 Sol (medium) | 76.3 |
| 21 | Qwen3.8 Max (0902) | 76.2 |
| 22 | GPT-6 Astra (Non-reasoning) | 76.2 |
| 23 | Kimi K3 (max) | 76.2 |
| 24 | Gemini 3.7 Flash (high) | 76.1 |
| 25 | GPT-6 Astra (xhigh) | 75.9 |
| 26 | Grok 4.6 (xhigh) | 75.9 |
| 27 | Muse Spark 1.3 (max) | 75.8 |
| 28 | GPT-6 Astra (low) | 75.7 |
| 29 | Claude Fable 5.1 (Adaptive Reasoning, Low Effort, Default Fallback) | 75.2 |
| 30 | GPT-5.5 (xhigh) | 74.9 |
| 31 | GLM-5.3 (max) | 74.8 |
| 32 | Grok 4.6 (medium) | 74.4 |
| 33 | Claude Opus 5 (Adaptive Reasoning, Medium Effort) | 74.3 |
| 34 | Claude Opus 4.8 (Adaptive Reasoning, Max Effort) | 74.3 |
| 35 | Gemini 3.8 Flash (medium) | 74.1 |
| 36 | Claude Opus 4.7 (Adaptive Reasoning, Max Effort) | 73.6 |
| 37 | Gemini 3.8 Flash (low) | 73.5 |
| 38 | Qwen3.8-Flash-Next | 73.1 |
| 39 | Grok 4.5 (high) | 72.4 |
| 40 | Muse Spark 1.2 (xhigh) | 72.2 |
| 41 | Kimi K3 (low) | 72.0 |
| 42 | Qwen3.8 2.4T A95B | 71.9 |
| 43 | Qwen3.8 Max | 71.8 |
| 44 | GPT-5.5 (high) | 71.6 |
| 45 | GLM 5.3 Flash | 71.5 |
| 46 | Gemini 3.7 Flash (medium) | 71.5 |
| 47 | Claude Sonnet 5 (Adaptive Reasoning, Max Effort) | 71.5 |
| 48 | GPT-5.5 (medium) | 71.5 |
| 49 | GPT-5.6 Luna (max) | 71.4 |
| 50 | Muse Spark 1.1 (xhigh) | 71.3 |
Quality, pricing, and speed benchmarks via Artificial Analysis. Human preference Elo via LMArena. Refreshed daily at 06:00 UTC · last sync .
About the metrics
Intelligence Index
AA
Artificial Analysis composite of reasoning, knowledge, coding, math.
LMArena Elo
LMArena
Human preference Elo from chatbot arena head-to-head votes.
Coding Index
AA
AA composite — LiveCodeBench, SciCode, etc.
Math Index
AA
AA composite — MATH-500, AIME, etc.
MMLU-Pro
AA
Multi-domain knowledge & reasoning. Harder than MMLU.
GPQA Diamond
AA
Graduate-level science Q&A. Strong reasoning signal.
LiveCodeBench
AA
Competitive programming, contamination-resistant.
MATH-500
AA
Subset of the MATH benchmark.
AIME
AA
American Invitational Mathematics Examination — olympiad math.
Humanity's Last Exam
AA
Frontier reasoning benchmark; most models still score low.
SciCode
AA
Scientific coding tasks.