SmophyAI

Published daily · Next update 02:00 UTC

Daily snapshot · immutable permalink

AI model usage rankings - August 17, 2026

As published on August 17, 2026, DeepSeek: DeepSeek V4 Flash 0423 led real OpenRouter usage share at 18.2%. Qwen: Qwen3.7 Max, MoonshotAI: Kimi K2.6, Anthropic: Claude Opus 4.6 (batch), Google: Gemini 3.1 Pro Preview (batch) entered the leaderboard for the first time. Anthropic: Claude Opus 4.7 (batch) moved 19 places up since August 16, 2026.

  1. #1 DeepSeek: DeepSeek V4 Flash 0423 - 18.2% usage share
  2. #2 Xiaomi: MiMo-V2.5 - 14.0% usage share
  3. #3 Tencent: Hy3 - 11.4% usage share
  4. #4 OpenAI: GPT-5.6 Luna (batch) - 7.9% usage share
  5. #5 Z.ai: GLM 5.2 (free) - 4.8% usage share
  6. #6 Google: Gemini 3.6 Flash (batch) - 4.1% usage share
  7. #7 NVIDIA: Nemotron 3 Ultra (free) - 3.9% usage share
  8. #8 DeepSeek: DeepSeek V4 Pro 0423 - 3.6% usage share
  9. #9 Claude Opus 5 (batch) - 2.0% usage share
  10. #10 MiniMax: MiniMax M3 (free) - 1.9% usage share
Change since August 16, 2026
#ModelUsage shareRank change
01DeepSeek: DeepSeek V4 Flash 042318.2%no change
02Xiaomi: MiMo-V2.514.0%+3 (+8.4pts)
03Tencent: Hy311.4%-1 (+1.1pts)
04OpenAI: GPT-5.6 Luna (batch)7.9%no change
05Z.ai: GLM 5.2 (free)4.8%+1 (-0.6pts)
06Google: Gemini 3.6 Flash (batch)4.1%+2 (-0.7pts)
07NVIDIA: Nemotron 3 Ultra (free)3.9%no change
08DeepSeek: DeepSeek V4 Pro 04233.6%+1 (-0.7pts)
09Claude Opus 5 (batch)2.0%-6 (-5.7pts)
10MiniMax: MiniMax M3 (free)1.9%+1 (-0.1pts)
11Poolside: Laguna S 2.1 (free)1.7%-1 (-0.7pts)
12Google: Gemini 3.7 Flash (batch)1.4%no change
13MoonshotAI: Kimi K31.4%no change
14Anthropic: Claude Sonnet 5 (batch)1.2%+3 (+0.3pts)
15OpenAI: GPT-5.6 Sol (batch)1.2%+6 (+0.4pts)
16OpenAI: GPT-5.6 Terra (batch)1.1%no change
17StepFun: Step 3.7 Flash1.0%-3 (-0.1pts)
18NVIDIA: Nemotron 3.5 Lightning (free)0.9%-3 (-0.1pts)
19Google: Gemini 3 Flash Preview (batch)0.9%-1 (0pts)
20Anthropic: Claude Sonnet 4.6 (batch)0.8%no change
21Anthropic: Claude Opus 4.7 (batch)0.8%+19 (+0.5pts)
22Google: Gemini 2.5 Flash Lite (batch)0.8%-3 (-0.1pts)
23SpaceXAI: Grok 4.60.7%+9 (+0.3pts)
24OpenAI: gpt-oss-120b0.6%+4 (+0.1pts)
25Xiaomi: MiMo-V2.5-Pro0.6%+1 (0pts)
26Upstage: Solar Pro 40.6%-4 (-0.1pts)
27OpenAI: GPT-5.6 Luna Pro (batch)0.6%-4 (0pts)
28OpenAI: GPT-5.5 (batch)0.5%-4 (-0.1pts)
29Anthropic: Claude Opus 4.8 (batch)0.5%+5 (+0.1pts)
30Google: Gemini 2.5 Flash (batch)0.5%-1 (0pts)
31Google: Gemini 3.1 Flash Lite (batch)0.5%+2 (+0.1pts)
32NVIDIA: Nemotron 3 Super (free)0.5%-2 (0pts)
33DeepSeek: DeepSeek V3.20.4%-8 (-0.1pts)
34Google: Gemma 4 26B A4B (free)0.4%-7 (-0.1pts)
35Google: Gemma 4 31B (free)0.4%-4 (-0.1pts)
36Anthropic: Claude Fable 5 (batch)0.4%+1 (+0.1pts)
37Qwen: Qwen3.8 Max0.3%+2 (0pts)
38Anthropic: Claude Haiku 4.5 (batch)0.3%+4 (0pts)
39OpenAI: GPT-4o-mini (batch)0.2%+4 (0pts)
40Qwen: Qwen3.7 Max New0.2%-
41OpenAI: GPT-5.4 (batch)0.2%+4 (0pts)
42Mistral: Mistral Nemo0.2%-6 (-0.1pts)
43MiniMax: MiniMax M2.7 (free)0.2%-5 (0pts)
44MoonshotAI: Kimi K2.6 New0.2%-
45Google: Gemini 3.5 Flash (batch)0.2%-4 (0pts)
46Anthropic: Claude Opus 4.6 (batch) New0.2%-
47Google: Gemini 3.1 Pro Preview (batch) New0.2%-

This page is a permanent, unchanging record of real OpenRouter usage share as it stood on August 17, 2026. Later corrections to the data pipeline appear as new dated snapshots - this page's figures never change.

Source: OpenRouter (openrouter.ai/rankings), as of August 17, 2026.

Benchmark scores: Artificial Analysis (artificialanalysis.ai) via OpenRouter (openrouter.ai/rankings).

Methodology: www.smophy.ai/benchmark/methodology

Token counts originate from each provider's own tokenizer and are not directly comparable across providers.