Arena: Anthropic Dominates Chat in August 2026

The August 2026 Chat ranking from Text Arena places Anthropic at the top, consistently dominating the leading positions. Meta, Alibaba, and Google also feature in the top 10, while OpenAI appears only in the 17th and 19th spots. These results are based on blind duels rated by an Elo score, with subcategories for Search, Vision, and Document complementing the analysis.

Arena assigns an Elo score after blind duels

Arena evaluates models across several categories, including Chat, and distinguishes three subcategories within it: Search, Vision, and Document Arena. The system relies on blind confrontations where two systems respond to the same prompt before a user selects the answer deemed most convincing. The votes contribute to an Elo score, inspired by chess, which fluctuates with each duel: defeating a higher-ranked opponent yields more points, winning against a lower-ranked model counts for little, and losing to the latter costs points.

In the specialties measured in August 2026, Claude Opus 4.6 Search ranks first in Search, ahead of GPT-5.5 Search and Fable 5. In Vision, Fable 5 takes the top spot, followed by Qwen 3.8 Max in second and Claude Opus 4.7 High in third. In Document Arena, the analysis of documents, including large ones, sees Anthropic taking the top six positions, with a leading trio composed of Claude Opus 5 High, Opus 4.6, and Opus 4.6 High.

Leading in August 2026, Anthropic claims six of the top seven spots

Following the overhaul in the summer of 2026, the August 2026 Chat ranking primarily highlights Anthropic. The top three positions are held by Claude Fable 5, Claude Opus 4.6 High, and Claude Opus 4.7 High. Muse Spark 1.2 xHigh, a model from Meta, slots in at fourth place, followed by Claude Opus 4.6, Opus 4.7, and Opus 5 High in fifth, sixth, and seventh places, respectively. Alibaba's Qwen 3.8-max ranks eighth, Google's Gemini 3.7 Flash High is ninth, and Claude Opus 5 Max rounds out the top 10.

The complete top 20 and OpenAI's discreet presence

OpenAI appears only twice in the top twenty of the Chat, with GPT-5.5 High in 17th and GPT-5.6 Sol xHigh in 19th. For the August 2026 edition of Text Arena, the ranking is as follows, from 1st to 20th place: Claude Fable 5; Claude Opus 4.6 High; Claude Opus 4.7 High; Muse Spark 1.2; Claude Opus 4.6; Claude Opus 4.7; Claude Opus 5 High; Qwen 3.8 Max; Gemini 3.7 Flash High; Claude Opus 5 Max; Muse Spark 1.1; Kimi K3 Max; Muse Spark; Gemini 3.1 Pro Preview; Gemini 3 Pro; Gemini 3.6 Flash High; GPT-5.5 High; Claude Opus 4.8 High; GPT-5.6 Sol xHigh; Gemini 3.5 Flash High.