Daily snapshot · immutable permalink
AI model usage rankings - September 7, 2026
As published on September 7, 2026, Tencent: Hy4 preview led real OpenRouter usage share at 17.4%. SpaceXAI: Grok 4.6, Meta: Muse Spark 1.3 entered the leaderboard for the first time. Xiaomi: MiMo-V2.5 moved 16 places up since September 6, 2026.
- #1 Tencent: Hy4 preview - 17.4% usage share
- #2 OpenAI: GPT-5.6 Luna (batch) - 13.4% usage share
- #3 DeepSeek: DeepSeek V4 Flash 0423 - 11.8% usage share
- #4 Z.ai: GLM 5.3 Flash (batch) - 9.6% usage share
- #5 MiniMax: MiniMax M3 - 4.2% usage share
- #6 Xiaomi: MiMo-V2.5 - 3.8% usage share
- #7 NVIDIA: Nemotron 3 Ultra (free) - 3.0% usage share
- #8 Z.ai: GLM 5.3 (batch) - 2.8% usage share
- #9 Tencent: Hy3 - 2.6% usage share
- #10 Z.ai: GLM 5.2 (free) - 1.9% usage share
| # | Model | Usage share | Rank change |
|---|---|---|---|
| 01 | Tencent: Hy4 preview | 17.4% | +2 (+5.1pts) |
| 02 | OpenAI: GPT-5.6 Luna (batch) | 13.4% | -1 (-1.9pts) |
| 03 | DeepSeek: DeepSeek V4 Flash 0423 | 11.8% | -1 (-3.2pts) |
| 04 | Z.ai: GLM 5.3 Flash (batch) | 9.6% | no change |
| 05 | MiniMax: MiniMax M3 | 4.2% | +2 (+1.5pts) |
| 06 | Xiaomi: MiMo-V2.5 | 3.8% | +16 (+3pts) |
| 07 | NVIDIA: Nemotron 3 Ultra (free) | 3.0% | -2 (-0.7pts) |
| 08 | Z.ai: GLM 5.3 (batch) | 2.8% | -2 (0pts) |
| 09 | Tencent: Hy3 | 2.6% | no change |
| 10 | Z.ai: GLM 5.2 (free) | 1.9% | -2 (-0.3pts) |
| 11 | OpenAI: GPT-5.6 Sol (batch) | 1.7% | +8 (+0.6pts) |
| 12 | DeepSeek: DeepSeek V4 Pro 0423 | 1.6% | -1 (-0.1pts) |
| 13 | Anthropic: Claude Opus 5 (batch) | 1.6% | -1 (-0.2pts) |
| 14 | Meta: Muse Spark 1.3 Contributor | 1.5% | -4 (-0.4pts) |
| 15 | Google: Gemini 3.8 Flash (batch) | 1.5% | -1 (0pts) |
| 16 | Upstage: Solar Pro 4 | 1.3% | -3 (-0.4pts) |
| 17 | Anthropic: Claude Sonnet 4.6 (batch) | 1.2% | -2 (-0.1pts) |
| 18 | MoonshotAI: Kimi K3 (batch) | 1.2% | -2 (-0.1pts) |
| 19 | Anthropic: Claude Sonnet 5 (batch) | 1.0% | +2 (0pts) |
| 20 | Poolside: Laguna S 2.1 (free) | 0.8% | -3 (-0.3pts) |
| 21 | NVIDIA: Nemotron 3.5 Lightning (free) | 0.7% | -1 (-0.4pts) |
| 22 | inclusionAI: Ling 3.0 Flash Fin (free) | 0.7% | +2 (-0.1pts) |
| 23 | Google: Gemini 3 Flash Preview (batch) | 0.6% | no change |
| 24 | Google: Gemini 3.7 Flash (batch) | 0.6% | +2 (-0.1pts) |
| 25 | MiniMax: MiniMax M2.7 | 0.6% | -7 (-0.5pts) |
| 26 | Google: Gemini 2.5 Flash Lite (batch) | 0.6% | -1 (-0.1pts) |
| 27 | DeepSeek: DeepSeek V4 Flash Vision Exp | 0.5% | no change |
| 28 | OpenAI: GPT-6 Astra (batch) | 0.5% | +1 (0pts) |
| 29 | OpenAI: GPT-5.6 Terra (batch) | 0.4% | +9 (+0.1pts) |
| 30 | Qwen: Qwen3.8 Max | 0.4% | +13 (+0.1pts) |
| 31 | DeepSeek: DeepSeek V3.2 | 0.4% | -3 (-0.1pts) |
| 32 | Google: Gemini 2.5 Flash (batch) | 0.3% | -1 (0pts) |
| 33 | OpenAI: GPT-5.6 Luna Pro (batch) | 0.3% | +7 (0pts) |
| 34 | Google: Gemini 3.1 Flash Lite (batch) | 0.3% | no change |
| 35 | OpenAI: gpt-oss-120b | 0.3% | -3 (-0.1pts) |
| 36 | Anthropic: Claude Opus 4.8 (batch) | 0.3% | +6 (0pts) |
| 37 | SpaceXAI: Grok 4.6 New | 0.3% | - |
| 38 | StepFun: Step 3.7 Flash | 0.3% | -2 (-0.1pts) |
| 39 | Thinking Machines: Inkling (free) | 0.3% | -4 (-0.1pts) |
| 40 | Google: Gemma 4 26B A4B (free) | 0.3% | -7 (-0.1pts) |
| 41 | Google: Gemma 4 31B (free) | 0.3% | -4 (-0.1pts) |
| 42 | Anthropic: Claude Fable 5.1 (batch) | 0.2% | -1 (0pts) |
| 43 | OpenAI: gpt-oss-20b (batch) | 0.2% | -4 (-0.1pts) |
| 44 | Meta: Muse Spark 1.3 New | 0.2% | - |
| 45 | Qwen: Qwen3.8 27B (free) | 0.2% | +1 (0pts) |
This page is a permanent, unchanging record of real OpenRouter usage share as it stood on September 7, 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 7, 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.