LMArena Text Elo leaderboard: which LLMs score highest
LMArena Text Elo (head-to-head chat preference (crowd votes)) currently has scores for 28 models. The top score is 1,525 by gemini-4-argon-high; the median model scores 1,478 and the lowest scores 1,468.
| # | Model | Organisation | LMArena Text Elo score | Cheapest input $/M | Source | Captured |
|---|---|---|---|---|---|---|
| 1 | gemini-4-argon-high | 1,525 | — | LMArena | 2026-10-05 | |
| 2 | Claude Fable 5.1 (max) | Anthropic | 1,501 | $10.00 | LMArena | 2026-10-05 |
| 3 | Gemini 3.8 Flash (high) | Google DeepMind | 1,495 | $0.750 | LMArena | 2026-10-05 |
| 4 | Muse Spark 1.2 (xhigh) | Meta AI | 1,494 | $1.25 | LMArena | 2026-10-05 |
| 5 | Muse Spark 1.3 (max) | Meta AI | 1,494 | $1.25 | LMArena | 2026-10-05 |
| 6 | Claude Opus 5 (max) | Anthropic | 1,489 | $5.00 | LMArena | 2026-10-05 |
| 7 | Muse Spark | Meta AI | 1,489 | — | LMArena | 2026-10-05 |
| 8 | Gemini 3.7 Flash (high) | Google DeepMind | 1,488 | $0.750 | LMArena | 2026-10-05 |
| 9 | Kimi K3 (max) | Moonshot | 1,488 | $1.29 | LMArena | 2026-10-05 |
| 10 | Gemini 3 Pro | Google DeepMind | 1,485 | $2.00 | LMArena | 2026-10-05 |
| 11 | Gemini 3.6 Flash (High) | 1,483 | $0.750 | LMArena | 2026-10-05 | |
| 12 | GPT-6.1 Sol (max) | OpenAI | 1,483 | $2.00 | LMArena | 2026-10-05 |
| 13 | mimo-v2.6-pro | — | 1,480 | $0.435 | LMArena | 2026-10-05 |
| 14 | GLM-5.3 (max) | Z.ai (Zhipu AI) | 1,478 | $0.070 | LMArena | 2026-10-05 |
| 15 | gemini-3.5-flash-high | 1,477 | $1.50 | LMArena | 2026-10-05 | |
| 16 | GPT-6 Astra (max) | OpenAI | 1,477 | $10.00 | LMArena | 2026-10-05 |
| 17 | glm-5.2-max | Z.ai | 1,476 | — | LMArena | 2026-10-05 |
| 18 | gpt-5.2-chat-latest-20260210 | OpenAI | 1,476 | — | LMArena | 2026-10-05 |
| 19 | grok-4.20-beta1 | xAI | 1,475 | — | LMArena | 2026-10-05 |
| 20 | qwen3.7-max-preview | Alibaba | 1,475 | $1.25 | LMArena | 2026-10-05 |
| 21 | claude-opus-4-5-20251101-high-32k | Anthropic | 1,474 | $5.00 | LMArena | 2026-10-05 |
| 22 | deepseek-v4.1-flash-max | DeepSeek | 1,474 | — | LMArena | 2026-10-05 |
| 23 | Gemini 3 Flash | Google DeepMind | 1,473 | $0.500 | LMArena | 2026-10-05 |
| 24 | GPT-5.5 Instant | OpenAI | 1,473 | — | LMArena | 2026-10-05 |
| 25 | grok-4.20-beta-0309-reasoning | xAI | 1,472 | $1.25 | LMArena | 2026-10-05 |
| 26 | grok-4.20-multi-agent-beta-0309 | xAI | 1,471 | $1.25 | LMArena | 2026-10-05 |
| 27 | ernie-5.1 | — | 1,468 | $0.563 | LMArena | 2026-10-05 |
| 28 | mimo-v2.5-pro | Xiaomi Corp | 1,468 | $0.435 | 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 Gemini 3.8 Flash (high) at $0.750 per million input tokens (score 1,495).
What LMArena Text Elo measures
LMArena (formerly Chatbot Arena) ranks models with Elo-style ratings computed from anonymous head-to-head votes by real users, so it measures human preference rather than accuracy on a fixed test.
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) and LMArena (third-party snapshot 2026-10-04). Scores were last captured 2026-10-05; the Captured column gives each row's date.