Ranked by LMArena Text - 22 models
Best AI models for chat
In board data through October 15, 2026, Gemini 4 Argon ranks first on the official LMArena Text (Text Arena Overall) leaderboard with 1524±9 (Preliminary), 15 points ahead of Muse Wonder (Max) (Meta) at 1509±11. Its list price is $2 input / $10 output per 1M tokens. A fixed rule, the lowest output price per 1M tokens among models with a registry price that score within 5 points of the top score, picks the leader itself. The highest-placed open-weight model is Kimi K3, ranked 15th with 1487±5.
Board data through , checked . How this is ranked
LMArena Text (Text Arena Overall) ranking
| Rank | Model | Provider | LMArena Text | Source | Input / 1M | Output / 1M | Context | Weights |
|---|---|---|---|---|---|---|---|---|
| 1 | Gemini 4 ArgonBoard: Gemini 4 Argon (High) (Google) | 1524±9 (Preliminary)Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 1 (rank spread 1-3); 4,904 votes; license: Proprietary | arena.ai | $2 | $10 | 1.05M | Closed | |
| 2 | Muse Wonder (Max) (Meta) | - | 1509±11Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 2 (rank spread 1-11); 3,247 votes; license: Proprietary | arena.ai | - | - | - | - |
| 3 | Claude Opus 5.5Board: Claude Opus 5.5 (High) (Anthropic) | Anthropic | 1508±7Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 3 (rank spread 1-8); 8,300 votes; license: Proprietary | arena.ai | $4 | $20 | 1M | Closed |
| 4 | Claude Fable 5Board: Claude Fable 5 (High) (Anthropic) | Anthropic | 1504±4Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 4 (rank spread 2-11); 42,004 votes; license: Proprietary | arena.ai | $10 | $50 | 1M | Closed |
| 5 | Claude Opus 4.6Board: Claude Opus 4.6 (High) (Anthropic)+1 more row on the board | Anthropic | 1503±3Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 5 (rank spread 2-11); 81,394 votes; license: Proprietary | arena.ai | $5 | $25 | 1M | Closed |
| 6 | Claude Fable 5.1Board: Claude Fable 5.1 (Max) (Anthropic) | Anthropic | 1502±6Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 6 (rank spread 2-12); 14,590 votes; license: Proprietary | arena.ai | $10 | $50 | 1M | Closed |
| 7 | Claude Opus 4.7Board: Claude Opus 4.7 (High) (Anthropic)+1 more row on the board | Anthropic | 1501±4Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 7 (rank spread 2-12); 65,116 votes; license: Proprietary | arena.ai | $5 | $25 | 1M | Closed |
| 8 | Gemini 3.8 FlashBoard: Gemini 3.8 Flash (High) (Google) | 1496±5 (Preliminary)Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 9 (rank spread 2-18); 37,454 votes; license: Proprietary | arena.ai | $0.75 | $3.75 | 1.05M | Closed | |
| 9 | Muse Spark 1.3 (Max) (Meta) | - | 1494±6Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 10 (rank spread 3-22); 17,561 votes; license: Proprietary | arena.ai | - | - | - | - |
| 10 | Muse Spark 1.2 (xHigh) (Meta) | - | 1493±7Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 12 (rank spread 3-25); 8,248 votes; license: Proprietary | arena.ai | - | - | - | - |
| 11 | Muse Spark 1.1 (Meta) | - | 1490±4Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 13 (rank spread 8-25); 42,810 votes; license: Proprietary | arena.ai | - | - | - | - |
| 11 | Claude Opus 5Board: Claude Opus 5 (Max) (Anthropic)+1 more row on the board | Anthropic | 1490±5Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 14 (rank spread 8-27); 32,495 votes; license: Proprietary | arena.ai | $5 | $25 | 1M | Closed |
| 13 | Muse Spark (Meta) | - | 1489±6Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 16 (rank spread 8-30); 14,150 votes; license: Proprietary | arena.ai | - | - | - | - |
| 14 | Gemini 3.7 Flash (High) (Google) | - | 1488±5 (Preliminary)Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 17 (rank spread 9-30); 27,232 votes; license: Proprietary | arena.ai | - | - | - | - |
| 15 | Kimi K3Board: Kimi K3 (Max) (Moonshot) | Moonshot AI | 1487±5Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 18 (rank spread 10-31); 32,795 votes; license: Kimi K3 license | arena.ai | $3 | $15 | 1.05M | Open |
| 15 | Gemini 3.1 Pro PreviewBoard: Gemini 3.1 Pro Preview (Google) | 1487±3Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 19 (rank spread 10-28); 127,649 votes; license: Proprietary | arena.ai | $2 | $12 | 1.05M | Closed | |
| 17 | Gemini 3 Pro (Google) | - | 1485±4Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 20 (rank spread 11-31); 42,072 votes; license: Proprietary | arena.ai | - | - | - | - |
| 17 | GPT-5.6 SolBoard: GPT-5.6 Sol (xHigh) (OpenAI) | OpenAI | 1485±4Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 21 (rank spread 11-33); 40,442 votes; license: Proprietary | arena.ai | $4 | $20 | 1.05M | Closed |
| 19 | GPT-6.1 SolBoard: GPT-6.1 Sol (Max) (OpenAI) | OpenAI | 1484±8Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 22 (rank spread 9-44); 6,320 votes; license: Proprietary | arena.ai | $2 | $10 | 1.05M | Closed |
| 20 | Gemini 3.6 FlashBoard: Gemini 3.6 Flash (High) (Google) | 1483±4Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 23 (rank spread 12-40); 36,122 votes; license: Proprietary | arena.ai | $0.75 | $3.75 | 1.05M | Closed | |
| 21 | Qwen3.8-MaxBoard: Qwen3.8-Max (Alibaba) | Alibaba Cloud (Qwen) | 1482±5Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 24 (rank spread 12-43); 27,766 votes; license: Proprietary | arena.ai | $2 | $6 | 1M | Closed |
| 22 | Claude Opus 4.8Board: Claude Opus 4.8 (High) (Anthropic) | Anthropic | 1481±4Official leaderboard - Text Arena Overall, default leaderboard view (8,834,211 votes, 416 models); rank 25 (rank spread 15-42); 64,100 votes; license: Proprietary | arena.ai | $5 | $25 | 1M | Closed |
Prices in USD per 1M tokens at the standard rate, from each provider's pricing page. A dash means the board name does not match a model in our registry.
Related
API capabilities
Providers of the top 3
Curated picks
FAQ
What is the best AI model for chat?
Gemini 4 Argon ranks first by LMArena Text (Text Arena Overall), board data through October 15, 2026, with 1524±9 (Preliminary) (official leaderboard). One benchmark is not a verdict: settings differ between rows, so test on your own workload before you switch.
What is the cheapest AI model for chat near the top of LMArena Text?
Taking every ranked model within 5 points of the top score that has a price in our registry, Gemini 4 Argon has the lowest output price: $2 input / $10 output per 1M tokens, with 1524±9 (Preliminary) by LMArena Text (Text Arena Overall), board data through October 15, 2026.
What is the best open-weight model for chat?
Kimi K3 is the highest-placed open-weight model by LMArena Text (Text Arena Overall), board data through October 15, 2026, ranked 15th with 1487±5.
How is this chat ranking built?
Rows follow the LMArena Text (Text Arena Overall) board's own order, checked on October 17, 2026; its newest entry is dated October 15, 2026, and a use case is only published while that date is within 120 days of the check. 3 board rows repeat a model already listed under a different agent, effort level or step budget; each model keeps its first, highest-placed row, in the board's own order. Prices, context windows and weights come from each provider's first-party pricing page in our model registry; a model the registry does not list shows a dash.
Newsletter
Get new benchmark results as they land
One email per week. New model releases, the leaderboard moves that matter and what they mean for what you build.