Elo · Cost · Speed
Model leaderboard
Every generated entry, aggregated per model variant and ranked by Elo rating from blind pairwise votes, alongside community votes, output volume, estimated cost per task, and measured generation time. Pick an arena, click a column to sort.
Wondering what these numbers say about the best AI for coding? Read the full breakdown →
| Model | |||||||
|---|---|---|---|---|---|---|---|
| Muse Spark 1.1 · Thinking | 1500? | 0-0-0 | 0 | 5 | ~10,413 | $0.04 | — |
| Fable 5 · low | 1500? | 0-0-0 | 0 | 34 | ~11,240 | $0.56 | 20m 18s |
| Fable 5 · medium | 1500? | 0-0-0 | 0 | 20 | ~13,622 | $0.68 | 26m 06s |
| Fable 5 · high | 1500? | 0-0-0 | 0 | 32 | ~13,137 | $0.66 | 32m 10s |
| Fable 5 · xhigh | 1500? | 0-0-0 | 0 | 32 | ~12,646 | $0.63 | 33m 36s |
| Fable 5 · max (post-ban) | 1500? | 0-0-0 | 0 | 19 | ~14,612 | $0.73 | 34m 34s |
| Fable 5 (pre-ban) | 1500? | 0-0-0 | 0 | 18 | ~15,339 | $0.77 | — |
| Opus 4.8 · low | 1500? | 0-0-0 | 0 | 40 | ~13,851 | $0.35 | — |
| Opus 4.8 · medium | 1500? | 0-0-0 | 0 | 40 | ~14,769 | $0.37 | — |
| Opus 4.8 · high | 1500? | 0-0-0 | 0 | 40 | ~16,804 | $0.42 | — |
| Opus 4.8 · xhigh | 1500? | 0-0-0 | 0 | 40 | ~16,503 | $0.41 | — |
| Opus 4.8 · max | 1500? | 0-0-0 | 0 | 40 | ~16,891 | $0.42 | — |
| Opus 5 · low | 1500? | 0-0-0 | 0 | 40 | ~15,777 | $0.39 | — |
| Opus 5 · medium | 1500? | 0-0-0 | 0 | 40 | ~16,001 | $0.40 | — |
| Opus 5 · high | 1500? | 0-0-0 | 0 | 40 | ~15,541 | $0.39 | — |
| Opus 5 · xhigh | 1500? | 0-0-0 | 0 | 40 | ~16,030 | $0.40 | — |
| Opus 5 · max | 1500? | 0-0-0 | 0 | 40 | ~16,374 | $0.41 | — |
| Opus 5 · ultracode | 1500? | 0-0-0 | 0 | 16 | ~29,278 | $0.73 | — |
| Sonnet 4.6 · low | 1500? | 0-0-0 | 0 | 40 | ~9,603 | $0.14 | — |
| Sonnet 4.6 · medium | 1500? | 0-0-0 | 0 | 40 | ~10,308 | $0.15 | — |
| Sonnet 4.6 · high | 1500? | 0-0-0 | 0 | 40 | ~11,062 | $0.17 | — |
| Sonnet 4.6 · max | 1500? | 0-0-0 | 0 | 40 | ~11,290 | $0.17 | — |
| Sonnet 5 · low | 1500? | 0-0-0 | 0 | 11 | ~10,154 | $0.15 | — |
| Sonnet 5 · medium | 1500? | 0-0-0 | 0 | 7 | ~13,008 | $0.20 | — |
| Sonnet 5 · high | 1500? | 0-0-0 | 0 | 10 | ~12,128 | $0.18 | — |
| Sonnet 5 · xhigh | 1500? | 0-0-0 | 0 | 8 | ~12,583 | $0.19 | — |
| Sonnet 5 · max | 1500? | 0-0-0 | 0 | 8 | ~13,260 | $0.20 | — |
| Haiku 4.5 | 1500? | 0-0-0 | 0 | 40 | ~7,235 | $0.04 | — |
| GLM-5.2 | 1500? | 0-0-0 | 0 | 19 | ~11,407 | $0.05 | — |
| GLM-5.2 · flash | 1500? | 0-0-0 | 0 | 40 | ~12,940 | $0.06 | — |
| GLM-5.1 | 1500? | 0-0-0 | 0 | 10 | ~12,764 | $0.06 | — |
| GLM-5 | 1500? | 0-0-0 | 0 | 10 | ~15,523 | $0.07 | — |
| GLM-5 Turbo | 1500? | 0-0-0 | 0 | 33 | ~8,632 | $0.04 | — |
| GLM-4.7 | 1500? | 0-0-0 | 0 | 40 | ~9,471 | $0.04 | — |
| GLM-4.6 | 1500? | 0-0-0 | 0 | 40 | ~8,976 | $0.04 | — |
| GLM-4.5 | 1500? | 0-0-0 | 0 | 40 | ~10,441 | $0.05 | — |
| GLM-4.5 Air | 1500? | 0-0-0 | 0 | 40 | ~8,936 | $0.04 | — |
| GPT-5.5 | 1500? | 0-0-0 | 0 | 40 | ~6,002 | $0.06 | — |
| Gemini 3 Flash | 1500? | 0-0-0 | 0 | 40 | ~7,137 | $0.02 | — |
| Kimi K2.7 Code | 1500? | 0-0-0 | 0 | 40 | ~7,868 | $0.03 | — |
| Qwen3.7 Plus | 1500? | 0-0-0 | 0 | 40 | ~7,373 | $0.01 | — |
| DeepSeek V4 Flash | 1500? | 0-0-0 | 0 | 40 | ~14,960 | $0.00 | — |
| MiniMax M3 | 1500? | 0-0-0 | 0 | 38 | ~6,447 | $0.01 | — |
| Kimi K2.6 | 1500? | 0-0-0 | 0 | 40 | ~11,664 | $0.05 | — |
| Kimi K3 | 1500? | 0-0-0 | 0 | 36 | ~8,522 | $0.03 | — |
| Qwen3.7 Max | 1500? | 0-0-0 | 0 | 40 | ~8,156 | $0.01 | — |
| Qwen3.8 Max | 1500? | 0-0-0 | 0 | 35 | ~12,513 | $0.02 | — |
| DeepSeek V4 Pro | 1500? | 0-0-0 | 0 | 40 | ~16,903 | $0.00 | — |
| MiniMax M2.7 | 1500? | 0-0-0 | 0 | 40 | ~10,658 | $0.01 | — |
| GPT-5.5 · low | 1500? | 0-0-0 | 0 | 40 | ~4,567 | $0.05 | — |
| GPT-5.5 · medium | 1500? | 0-0-0 | 0 | 40 | ~5,031 | $0.05 | — |
| GPT-5.5 · high | 1500? | 0-0-0 | 0 | 40 | ~6,266 | $0.06 | — |
| GPT-5.5 · xhigh | 1500? | 0-0-0 | 0 | 40 | ~8,413 | $0.08 | — |
| GPT-5.6 Luna · low | 1500? | 0-0-0 | 0 | 40 | ~3,569 | $0.04 | — |
| GPT-5.6 Luna · max | 1500? | 0-0-0 | 0 | 40 | ~12,188 | $0.12 | — |
| GPT-5.6 Terra · low | 1500? | 0-0-0 | 0 | 40 | ~3,904 | $0.04 | — |
| GPT-5.6 Terra · max | 1500? | 0-0-0 | 0 | 40 | ~9,746 | $0.10 | — |
| GPT-5.6 Sol · low | 1500? | 0-0-0 | 0 | 40 | ~5,329 | $0.05 | — |
| GPT-5.6 Sol · max | 1500? | 0-0-0 | 0 | 40 | ~12,092 | $0.12 | — |
| GPT-5.6 Luna · ultra | 1500? | 0-0-0 | 0 | 40 | ~16,692 | $0.17 | — |
| GPT-5.6 Terra · ultra | 1500? | 0-0-0 | 0 | 40 | ~18,160 | $0.18 | — |
| GPT-5.6 Sol · ultra | 1500? | 0-0-0 | 0 | 40 | ~20,575 | $0.21 | — |
| GPT-5.6 Luna Pro | 1500? | 0-0-0 | 0 | 1 | ~6,246 | $0.06 | — |
| GPT-5.6 Luna · medium | 1500? | 0-0-0 | 0 | 40 | ~4,650 | $0.05 | — |
| GPT-5.6 Luna · high | 1500? | 0-0-0 | 0 | 40 | ~8,501 | $0.09 | — |
| GPT-5.6 Terra · medium | 1500? | 0-0-0 | 0 | 40 | ~4,105 | $0.04 | — |
| GPT-5.6 Terra · high | 1500? | 0-0-0 | 0 | 40 | ~4,950 | $0.05 | — |
| GPT-5.6 Sol · medium | 1500? | 0-0-0 | 0 | 40 | ~7,536 | $0.08 | — |
| GPT-5.6 Sol · high | 1500? | 0-0-0 | 0 | 40 | ~10,704 | $0.11 | — |
| Mistral Large 2512 | 1500? | 0-0-0 | 0 | 40 | ~11,170 | $0.02 | — |
| Mistral Medium 3.5 | 1500? | 0-0-0 | 0 | 40 | ~9,347 | $0.02 | — |
| Mistral Medium 3 (2508) | 1500? | 0-0-0 | 0 | 39 | ~17,512 | $0.04 | — |
| Mistral Medium (2505) | 1500? | 0-0-0 | 0 | 39 | ~8,534 | $0.02 | — |
| Mistral Small 2603 | 1500? | 0-0-0 | 0 | 40 | ~6,386 | $0.01 | — |
| Mistral Small 2506 | 1500? | 0-0-0 | 0 | 39 | ~5,542 | $0.01 | — |
| Codestral 2508 | 1500? | 0-0-0 | 0 | 40 | ~7,946 | $0.02 | — |
| Devstral 2512 | 1500? | 0-0-0 | 0 | 39 | ~10,327 | $0.02 | — |
| Magistral Medium · reasoning | 1500? | 0-0-0 | 0 | 10 | ~6,892 | $0.01 | — |
| Magistral Medium · no-reason | 1500? | 0-0-0 | 0 | 39 | ~8,877 | $0.02 | — |
| Magistral Small · reasoning | 1500? | 0-0-0 | 0 | 40 | ~6,119 | $0.01 | — |
| Magistral Small · no-reason | 1500? | 0-0-0 | 0 | 40 | ~5,536 | $0.01 | — |
| Ministral 14B | 1500? | 0-0-0 | 0 | 39 | ~14,725 | $0.03 | — |
| Ministral 8B | 1500? | 0-0-0 | 0 | 37 | ~12,817 | $0.03 | — |
| Ministral 3B | 1500? | 0-0-0 | 0 | 38 | ~11,777 | $0.02 | — |
| Mistral Nemo | 1500? | 0-0-0 | 0 | 40 | ~1,707 | $0.00 | — |
| Grok 4.3 | 1500? | 0-0-0 | 0 | 40 | ~3,338 | $0.01 | — |
| Grok 4.20 · non-reasoning | 1500? | 0-0-0 | 0 | 40 | ~11,924 | $0.03 | — |
| Grok 4.20 · reasoning | 1500? | 0-0-0 | 0 | 39 | ~11,550 | $0.03 | — |
| Grok 4.5 · high | 1500? | 0-0-0 | 0 | 39 | ~11,422 | $0.03 | — |
| Grok 4.5 · low | 1500? | 0-0-0 | 0 | 38 | ~8,915 | $0.02 | — |
| GLM-5.2 · OR fp8 | 1500? | 0-0-0 | 0 | 18 | ~9,913 | $0.04 | — |
| GLM-5.1 · OR fp8 | 1500? | 0-0-0 | 0 | 18 | ~10,210 | $0.04 | — |
| GLM-5 · OR fp8 | 1500? | 0-0-0 | 0 | 1 | ~12,716 | $0.06 | — |
| GLM-4.7 · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~9,770 | $0.04 | — |
| GLM-4.7 Flash · OR fp8 | 1500? | 0-0-0 | 0 | 18 | ~8,716 | $0.04 | — |
| GLM-4.5 Air · OR fp8 | 1500? | 0-0-0 | 0 | 19 | ~11,441 | $0.05 | — |
| DeepSeek V4 Flash · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~16,445 | $0.00 | — |
| DeepSeek V4 Pro · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~15,988 | $0.00 | — |
| DeepSeek V3.1 · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~7,119 | $0.00 | — |
| DeepSeek R1-0528 · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~11,440 | $0.00 | — |
| Kimi K2.5 · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~8,958 | $0.04 | — |
| Kimi K2.6 · OR fp8 | 1500? | 0-0-0 | 0 | 19 | ~11,931 | $0.05 | — |
| Kimi K2.7 Code · OR fp8 | 1500? | 0-0-0 | 0 | 18 | ~8,965 | $0.04 | — |
| Kimi K2-0905 · OR fp8 | 1500? | 0-0-0 | 0 | 4 | ~10,882 | $0.04 | — |
| MiniMax M2.7 · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~10,883 | $0.01 | — |
| MiniMax M2.5 · OR fp8 | 1500? | 0-0-0 | 0 | 4 | ~13,211 | $0.02 | — |
| MiniMax M3 · OR fp8 | 1500? | 0-0-0 | 0 | 7 | ~13,385 | $0.02 | — |
| MiniMax M2.1 · OR fp8 | 1500? | 0-0-0 | 0 | 4 | ~11,294 | $0.01 | — |
| MiniMax M2 · OR fp8 | 1500? | 0-0-0 | 0 | 4 | ~11,771 | $0.01 | — |
| Qwen3.6 27B · OR fp8 | 1500? | 0-0-0 | 0 | 19 | ~9,642 | $0.02 | — |
| Qwen3.6 35B-A3B · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~10,059 | $0.02 | — |
| Qwen3.5 122B-A10B · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~7,611 | $0.01 | — |
| Qwen3.5 397B-A17B · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~8,165 | $0.01 | — |
| Qwen3.5 27B · OR fp8 | 1500? | 0-0-0 | 0 | 19 | ~7,643 | $0.01 | — |
| Qwen3.5 9B · OR fp8 | 1500? | 0-0-0 | 0 | 19 | ~6,802 | $0.01 | — |
| Qwen3.5 35B-A3B · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~8,688 | $0.01 | — |
| Qwen3 Coder Next · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~9,954 | $0.02 | — |
| Qwen3 Coder 480B · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~7,619 | $0.01 | — |
| Qwen3 Coder 30B · OR fp8 | 1500? | 0-0-0 | 0 | 17 | ~7,835 | $0.01 | — |
| Qwen3 235B-2507 · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~11,653 | $0.02 | — |
| Qwen3 32B · OR fp8 | 1500? | 0-0-0 | 0 | 18 | ~5,788 | $0.01 | — |
| Qwen3 30B-A3B · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~11,637 | $0.02 | — |
| gpt-oss-120b · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~3,767 | $0.00 | — |
| gpt-oss-20b · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~2,798 | $0.00 | — |
| Gemma 4 26B-A4B · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~5,836 | $0.00 | — |
| Gemma 4 31B · OR fp8 | 1500? | 0-0-0 | 0 | 18 | ~5,571 | $0.00 | — |
| Gemma 3 27B · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~2,954 | $0.00 | — |
| Llama 4 Maverick · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~3,194 | $0.00 | — |
| Llama 4 Scout · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~2,376 | $0.00 | — |
| Llama 3.3 70B · OR fp8 | 1500? | 0-0-0 | 0 | 20 | ~2,922 | $0.00 | — |
| Llama 3.1 8B · OR fp8 | 1500? | 0-0-0 | 0 | 15 | ~5,576 | $0.00 | — |
| MiMo V2.5 · OR fp8 | 1500? | 0-0-0 | 0 | 14 | ~9,344 | $0.01 | — |
| MiMo V2.5 Pro · OR fp8 | 1500? | 0-0-0 | 0 | 5 | ~11,484 | $0.01 | — |
| Step 3.7 Flash · OR fp8 | 1500? | 0-0-0 | 0 | 18 | ~10,897 | $0.02 | — |
| Nemotron 3 Super 120B · OR fp8 | 1500? | 0-0-0 | 0 | 18 | ~4,905 | $0.01 | — |
| Nemotron 3 Nano 30B · OR fp8 | 1500? | 0-0-0 | 0 | 17 | ~3,643 | $0.01 | — |
| Mistral Small 3.2 · OR fp8 | 1500? | 0-0-0 | 0 | 19 | ~5,856 | $0.01 | — |
| Ministral 14B · OR fp8 | 1500? | 0-0-0 | 0 | 16 | ~13,755 | $0.03 | — |
| Hunyuan A13B · OR fp8 | 1500? | 0-0-0 | 0 | 15 | ~3,326 | $0.00 | — |
Elo ratings replay the public pairwise match log: every blind vote records the models that were on screen, the winner beats each shown opponent (ties count 0.5), starting at 1500 with K 32 for a variant's first 30 matches and 16 after. One vote carries the same total weight no matter how many panes were compared. Variants with no rated matches yet display the 1500 baseline. Ratings are per arena and started fresh on 2026-07-10; earlier votes lacked opponent data and stay in the raw vote counts only. Token counts are estimated from artifact file size (chars ÷ 4). Cost = estimated output tokens × the model's published per-1M output price; input and reasoning tokens aren't counted, so true cost is higher, especially at high thinking effort. Generation time is wall-clock and only recorded for runs generated after 2026-07-02; it includes queue and throttling waits. This page compares like-for-like one-shot generations, not lab benchmarks.