LLM Leaderboard
Ranked by SciCode · source AA
Scientific coding tasks.
#1 by SciCode
Claude Opus 5.5 (Max, Default Fallback)
66.9%
Anthropic · in $4.00 / out $20.0 per 1M · 96 tok/s
| # | Model | SciCode |
|---|---|---|
| 1 | Claude Opus 5.5 (Max, Default Fallback) | 66.9% |
| 2 | Claude Opus 5.5 (Xhigh, Default Fallback) | 65.0% |
| 3 | Claude Fable 5.1 (Max, Default Fallback) | 63.1% |
| 4 | Muse Wonder Full (Max) | 62.6% |
| 5 | Gemini 4 Argon (High) | 61.8% |
| 6 | Claude Sonnet 5.5 (Max, Default Fallback) | 61.0% |
| 7 | Claude Fable 5 (Max, Opus 4.8 Fallback) | 61.0% |
| 8 | Claude Fable 5.1 (Xhigh, Default Fallback) | 60.9% |
| 9 | MiMo-V2.6-Pro | 60.9% |
| 10 | Kimi K3.1 Pro | 60.8% |
| 11 | Claude Opus 5.5 (High, Default Fallback) | 60.4% |
| 12 | Muse Wonder (Max) | 60.3% |
| 13 | Gemini 3.7 Flash (Medium) | 59.8% |
| 14 | Muse Spark 1.3 (Xhigh) | 59.7% |
| 15 | Kimi K3 (Max) | 59.5% |
| 16 | Claude Opus 5.5 (Medium, Default Fallback) | 59.3% |
| 17 | GLM-5.3 (Max) | 59.0% |
| 18 | Step 5 Preview | 58.9% |
| 19 | Muse Spark 1.3 (Max) | 58.8% |
| 20 | Muse Spark 1.1 (Xhigh) | 58.8% |
| 21 | Claude Fable 5.1 (High, Default Fallback) | 58.7% |
| 22 | Gemini 3.1 Pro Preview | 58.7% |
| 23 | Claude Opus 5.5 (Low, Default Fallback) | 58.6% |
| 24 | Muse Spark Open (Xhigh) | 58.4% |
| 25 | Muse Spark 1.3 (max) | 58.3% |
| 26 | Grok 4.7 (High) | 57.8% |
| 27 | GPT-5.6 Sol (High) | 57.8% |
| 28 | GPT-6 Sol (Max) | 57.6% |
| 29 | Grok 4.7 (Xhigh) | 57.4% |
| 30 | Muse Spark 1.2 (Xhigh) | 57.4% |
| 31 | GPT-5.6 Sol (Medium) | 57.4% |
| 32 | Claude Sonnet 5.5 (Xhigh, Default Fallback) | 57.3% |
| 33 | Gemini 3.7 Flash (High) | 57.2% |
| 34 | GPT-5.6 Sol (Xhigh) | 57.1% |
| 35 | GPT-5.6 Sol (Max) | 57.1% |
| 36 | Ling 3.1 Flash 1019 | 56.8% |
| 37 | Claude Fable 5.1 (Low, Default Fallback) | 56.7% |
| 38 | Gemini 3.8 Flash (High) | 56.6% |
| 39 | GPT-6 Astra (Max) | 56.5% |
| 40 | Grok 4.6 (High) | 56.5% |
| 41 | Claude Fable 5.1 (Medium, Default Fallback) | 56.4% |
| 42 | GPT-5.6 Sol (Low) | 56.4% |
| 43 | Claude Opus 5 (Max) | 56.4% |
| 44 | GPT-5.5 (High) | 56.1% |
| 45 | DeepSeek V4.1 Flash 1017 (Max) | 55.9% |
| 46 | Grok 4.6 (Medium) | 55.9% |
| 47 | GPT-6.1 Sol (High) | 55.8% |
| 48 | GPT-5.5 (Xhigh) | 55.8% |
| 49 | GPT-6 Astra (Xhigh) | 55.7% |
| 50 | GPT-6.1 Sol (Xhigh) | 55.7% |
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.