SmophyAI

Published daily · Next update 02:00 UTC

Daily snapshot · immutable permalink

AI model usage rankings - August 15, 2026

As published on August 15, 2026, DeepSeek: DeepSeek V4 Flash 0423 led real OpenRouter usage share at 20.8%. Nex AGI: Nex-N2-Mini, Qwen: Qwen3.7 Flash, Poolside: Laguna XS 2.1 (free) entered the leaderboard for the first time. Anthropic: Claude Opus 4.8 (batch) moved 17 places down since August 14, 2026.

  1. #1 DeepSeek: DeepSeek V4 Flash 0423 - 20.8% usage share
  2. #2 Tencent: Hy3 - 11.7% usage share
  3. #3 OpenAI: GPT-5.6 Luna (batch) - 6.8% usage share
  4. #4 Claude Opus 5 (batch) - 5.5% usage share
  5. #5 Z.ai: GLM 5.2 (free) - 5.5% usage share
  6. #6 DeepSeek: DeepSeek V4 Pro 0423 - 5.1% usage share
  7. #7 Xiaomi: MiMo-V2.5 - 4.7% usage share
  8. #8 NVIDIA: Nemotron 3 Ultra (free) - 4.2% usage share
  9. #9 Poolside: Laguna S 2.1 (free) - 2.5% usage share
  10. #10 MiniMax: MiniMax M3 (free) - 2.2% usage share
Change since August 14, 2026
#ModelUsage shareRank change
01DeepSeek: DeepSeek V4 Flash 042320.8%no change
02Tencent: Hy311.7%no change
03OpenAI: GPT-5.6 Luna (batch)6.8%+1 (+0.3pts)
04Claude Opus 5 (batch)5.5%+5 (+2.9pts)
05Z.ai: GLM 5.2 (free)5.5%-2 (-2.4pts)
06DeepSeek: DeepSeek V4 Pro 04235.1%-1 (+0.5pts)
07Xiaomi: MiMo-V2.54.7%no change
08NVIDIA: Nemotron 3 Ultra (free)4.2%no change
09Poolside: Laguna S 2.1 (free)2.5%+1 (+0.4pts)
10MiniMax: MiniMax M3 (free)2.2%+1 (+0.2pts)
11MoonshotAI: Kimi K31.5%+1 (-0.4pts)
12Google: Gemini 3.6 Flash (batch)1.4%-6 (-3.1pts)
13Google: Gemini 3.7 Flash (batch)1.3%+6 (+0.3pts)
14Anthropic: Claude Opus 4.7 (batch)1.3%+1 (+0.1pts)
15NVIDIA: Nemotron 3.5 Lightning (free)1.3%+3 (+0.2pts)
16StepFun: Step 3.7 Flash1.2%no change
17OpenAI: GPT-5.6 Terra (batch)1.2%-3 (-0.1pts)
18Anthropic: Claude Sonnet 5 (batch)1.1%-5 (-0.4pts)
19Google: Gemini 2.5 Flash Lite (batch)0.9%+3 (0pts)
20Google: Gemini 3 Flash Preview (batch)0.9%no change
21OpenAI: GPT-5.6 Sol (batch)0.9%+2 (0pts)
22Anthropic: Claude Sonnet 4.6 (batch)0.8%-1 (-0.1pts)
23OpenAI: GPT-5.6 Luna Pro (batch)0.7%+1 (-0.1pts)
24Upstage: Solar Pro 40.7%+9 (+0.2pts)
25DeepSeek: DeepSeek V3.20.6%+5 (+0.1pts)
26Xiaomi: MiMo-V2.5-Pro0.6%no change
27SpaceXAI: Grok 4.60.6%-2 (-0.2pts)
28NVIDIA: Nemotron 3 Super (free)0.5%+4 (+0.1pts)
29Google: Gemini 2.5 Flash (batch)0.5%-1 (0pts)
30OpenAI: gpt-oss-120b0.5%-1 (0pts)
31Google: Gemma 4 31B (free)0.5%no change
32Google: Gemma 4 26B A4B (free)0.5%+2 (+0.1pts)
33Google: Gemini 3.1 Flash Lite (batch)0.5%-6 (-0.1pts)
34Anthropic: Claude Opus 4.8 (batch)0.3%-17 (-0.7pts)
35Anthropic: Claude Fable 5 (batch)0.3%+4 (0pts)
36Qwen: Qwen3.8 Max0.3%no change
37Qwen: Qwen3.7 Max0.3%+10 (+0.1pts)
38Mistral: Mistral Nemo0.3%+8 (+0.1pts)
39Nex AGI: Nex-N2-Mini New0.3%-
40Anthropic: Claude Haiku 4.5 (batch)0.3%+1 (0pts)
41MiniMax: MiniMax M2.7 (free)0.3%+3 (0pts)
42OpenAI: GPT-5.4 (batch)0.3%+3 (0pts)
43OpenAI: GPT-4o-mini (batch)0.3%-1 (0pts)
44Qwen: Qwen3.7 Flash New0.3%-
45Anthropic: Claude Opus 4.6 (batch)0.3%-10 (-0.2pts)
46Google: Gemini 3.5 Flash (batch)0.2%-3 (0pts)
47Poolside: Laguna XS 2.1 (free) New0.2%-

This page is a permanent, unchanging record of real OpenRouter usage share as it stood on August 15, 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 15, 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.