LMArena Code Elo leaderboard: which LLMs score highest
LMArena Code Elo (crowd-voted coding answers) currently has scores for 29 models. The top score is 1,815 by Claude Opus 5.5 (max); the median model scores 1,620 and the lowest scores 1,512.
| # | Model | Organisation | LMArena Code Elo score | Cheapest input $/M | Source | Captured |
|---|---|---|---|---|---|---|
| 1 | Claude Opus 5.5 (max) | Anthropic | 1,815 | $4.00 | LMArena | 2026-10-05 |
| 2 | GPT-6 Astra (max) | OpenAI | 1,788 | $10.00 | LMArena | 2026-10-05 |
| 3 | GPT-6.1 Sol (max) | OpenAI | 1,758 | $2.00 | LMArena | 2026-10-05 |
| 4 | Claude Fable 5.1 (max) | Anthropic | 1,749 | $10.00 | LMArena | 2026-10-05 |
| 5 | Claude Opus 5 (max) | Anthropic | 1,695 | $5.00 | LMArena | 2026-10-05 |
| 6 | GPT-6 Sol (max) | OpenAI | 1,689 | $2.00 | LMArena | 2026-10-05 |
| 7 | gemini-4-argon-high | 1,680 | — | LMArena | 2026-10-05 | |
| 8 | Kimi K3 (max) | Moonshot | 1,658 | $1.29 | LMArena | 2026-10-05 |
| 9 | Muse Spark 1.3 (max) | Meta AI | 1,657 | $1.25 | LMArena | 2026-10-05 |
| 10 | Grok 4.7 (xhigh) | xAI | 1,638 | $2.00 | LMArena | 2026-10-05 |
| 11 | Qwen3.8-Flash-Next | Qwen | 1,637 | — | LMArena | 2026-10-05 |
| 12 | Hy4-preview | tencent | 1,633 | $0.751 | LMArena | 2026-10-05 |
| 13 | GLM-5.3 (max) | Z.ai (Zhipu AI) | 1,623 | $0.070 | LMArena | 2026-10-05 |
| 14 | deepseek-v4.1-flash-max | DeepSeek | 1,620 | — | LMArena | 2026-10-05 |
| 15 | gpt-5.6-sol-xhigh (codex-harness) | OpenAI | 1,620 | $2.00 | LMArena | 2026-10-05 |
| 16 | mimo-v2.6-pro | — | 1,618 | $0.435 | LMArena | 2026-10-05 |
| 17 | glm-5.2-max | Z.ai | 1,605 | — | LMArena | 2026-10-05 |
| 18 | Gemini 3.7 Flash (high) | Google DeepMind | 1,592 | $0.750 | LMArena | 2026-10-05 |
| 19 | deepseek-v4-pro-high-20260813 | DeepSeek | 1,583 | — | LMArena | 2026-10-05 |
| 20 | Gemini 3.8 Flash (high) | Google DeepMind | 1,583 | $0.750 | LMArena | 2026-10-05 |
| 21 | GPT-6 Luna (max) | OpenAI | 1,579 | $0.100 | LMArena | 2026-10-05 |
| 22 | step-5-preview-high | — | 1,570 | — | LMArena | 2026-10-05 |
| 23 | Gemini 3.6 Flash (High) | 1,536 | $0.750 | LMArena | 2026-10-05 | |
| 24 | Muse Spark 1.2 (xhigh) | Meta AI | 1,532 | $1.25 | LMArena | 2026-10-05 |
| 25 | gpt-5.6-terra-xhigh (codex-harness) | OpenAI | 1,519 | $2.00 | LMArena | 2026-10-05 |
| 26 | gpt-5.6-luna-xhigh (codex-harness) | OpenAI | 1,518 | $0.200 | LMArena | 2026-10-05 |
| 27 | seed-2.1-pro-preview | ByteDance | 1,517 | — | LMArena | 2026-10-05 |
| 28 | qwen3.7-max-20260517 | Alibaba | 1,515 | — | LMArena | 2026-10-05 |
| 29 | gpt-5.5-xhigh (codex-harness) | OpenAI | 1,512 | $5.00 | LMArena | 2026-10-05 |
Compare all benchmarks side by side on the live leaderboard →
Among the ten highest scorers, the cheapest listed API price belongs to Muse Spark 1.3 (max) at $1.25 per million input tokens (score 1,657).
What LMArena Code Elo measures
This is LMArena's coding category: Elo-style ratings from anonymous head-to-head votes on coding answers.
How to read these scores
Scores are Elo-style ratings (higher is better). Only differences between models mean anything, and small gaps may sit inside the source's own margin of error. Where a model is listed at several reasoning efforts, only its highest-effort row is shown. See the methodology for how rows are chosen.
Where the data comes from
LMArena (third-party snapshot 2026-10-05). Scores were last captured 2026-10-05; the Captured column gives each row's date.