SmophyAI

Published daily · Next update 02:00 UTC

Daily snapshot · immutable permalink

AI model usage rankings - September 4, 2026

As published on September 4, 2026, DeepSeek: DeepSeek V4 Flash 0423 led real OpenRouter usage share at 16.4%. Thinking Machines: Inkling (free), OpenAI: GPT-4o-mini (batch) entered the leaderboard for the first time. Anthropic: Claude Opus 4.8 (batch) moved 12 places up since September 3, 2026.

  1. #1 DeepSeek: DeepSeek V4 Flash 0423 - 16.4% usage share
  2. #2 Tencent: Hy4 preview - 16.2% usage share
  3. #3 Z.ai: GLM 5.3 Flash (batch) - 10.5% usage share
  4. #4 OpenAI: GPT-5.6 Luna (batch) - 7.6% usage share
  5. #5 MiniMax: MiniMax M3 - 6.2% usage share
  6. #6 Z.ai: GLM 5.3 (batch) - 3.3% usage share
  7. #7 NVIDIA: Nemotron 3 Ultra (free) - 2.9% usage share
  8. #8 Tencent: Hy3 - 2.8% usage share
  9. #9 Z.ai: GLM 5.2 (free) - 2.2% usage share
  10. #10 Anthropic: Claude Opus 5 (batch) - 1.9% usage share
Change since September 3, 2026
#ModelUsage shareRank change
01DeepSeek: DeepSeek V4 Flash 042316.4%+1 (+1.9pts)
02Tencent: Hy4 preview16.2%-1 (+0.1pts)
03Z.ai: GLM 5.3 Flash (batch)10.5%no change
04OpenAI: GPT-5.6 Luna (batch)7.6%no change
05MiniMax: MiniMax M36.2%no change
06Z.ai: GLM 5.3 (batch)3.3%+1 (+0.2pts)
07NVIDIA: Nemotron 3 Ultra (free)2.9%+1 (+0.1pts)
08Tencent: Hy32.8%-2 (-1.3pts)
09Z.ai: GLM 5.2 (free)2.2%+1 (+0.1pts)
10Anthropic: Claude Opus 5 (batch)1.9%+5 (+0.3pts)
11MoonshotAI: Kimi K3 (batch)1.8%+2 (-0.1pts)
12Google: Gemini 3.8 Flash (batch)1.7%+4 (+0.3pts)
13DeepSeek: DeepSeek V4 Pro 04231.7%-1 (-0.2pts)
14OpenAI: GPT-5.6 Sol (batch)1.4%-3 (-0.7pts)
15Xiaomi: MiMo-V2.51.3%-1 (-0.4pts)
16Anthropic: Claude Sonnet 5 (batch)1.2%+1 (+0.1pts)
17Google: Gemini 3.7 Flash (batch)1.2%-8 (-1pts)
18Anthropic: Claude Sonnet 4.6 (batch)1.2%no change
19Poolside: Laguna S 2.1 (free)1.1%no change
20Meta: Muse Spark 1.3 Contributor1.0%+2 (+0.2pts)
21NVIDIA: Nemotron 3.5 Lightning (free)0.9%-1 (+0.1pts)
22Upstage: Solar Pro 40.9%-1 (+0.1pts)
23inclusionAI: Ling 3.0 Flash Fin (free)0.7%+1 (+0.1pts)
24Google: Gemini 3 Flash Preview (batch)0.7%-1 (0pts)
25Google: Gemini 2.5 Flash Lite (batch)0.6%no change
26MiniMax: MiniMax M2.70.6%+1 (+0.1pts)
27OpenAI: GPT-5.6 Terra (batch)0.5%-1 (0pts)
28Anthropic: Claude Opus 4.8 (batch)0.4%+12 (+0.1pts)
29DeepSeek: DeepSeek V4 Flash Vision Exp0.4%+3 (0pts)
30Google: Gemini 2.5 Flash (batch)0.4%-2 (0pts)
31Anthropic: Claude Fable 5.1 (batch)0.4%-2 (0pts)
32SpaceXAI: Grok 4.60.4%+9 (+0.1pts)
33Google: Gemini 3.1 Flash Lite (batch)0.4%-2 (0pts)
34OpenAI: gpt-oss-120b0.4%no change
35DeepSeek: DeepSeek V3.20.3%-2 (0pts)
36StepFun: Step 3.7 Flash0.3%-6 (-0.1pts)
37NVIDIA: Nemotron 3 Super (free)0.3%+1 (0pts)
38Google: Gemma 4 26B A4B (free)0.3%-1 (0pts)
39OpenAI: GPT-5.6 Luna Pro (batch)0.3%-3 (0pts)
40Qwen: Qwen3.8 Max0.3%+4 (+0.1pts)
41Google: Gemma 4 31B (free)0.3%-2 (0pts)
42OpenAI: gpt-oss-20b (batch)0.3%-7 (-0.1pts)
43Xiaomi: MiMo-V2.5-Pro0.2%no change
44Thinking Machines: Inkling (free) New0.2%-
45Qwen: Qwen3.8 27B (free)0.2%+2 (0pts)
46OpenAI: GPT-4o-mini (batch) New0.2%-
47Anthropic: Claude Haiku 4.5 (batch)0.2%-2 (0pts)

This page is a permanent, unchanging record of real OpenRouter usage share as it stood on September 4, 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 September 4, 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.