Themed ranking
Best for agentic workloads
Ranked by real agentic-task usage share, weighted across five real OpenRouter agent tags (tool dispatch, workflow execution, planning, and more).
As of September 23, 2026, Z.ai: GLM 5.3 Flash (batch) ranks #1 for best for agentic workloads at agentic traffic share 19.3%.
- #1 Z.ai: GLM 5.3 Flash (batch) - Agentic traffic share 19.3%
- #2 DeepSeek: DeepSeek V4 Flash 0423 - Agentic traffic share 18.0%
- #3 DeepSeek: DeepSeek V4.1 Flash (batch) - Agentic traffic share 16.4%
- #4 OpenAI: GPT-5.6 Luna (batch) - Agentic traffic share 14.6%
- #5 Tencent: Hy4 preview - Agentic traffic share 10.8%
- #6 Tencent: Hy3 - Agentic traffic share 5.3%
- #7 Xiaomi: MiMo-V2.5 - Agentic traffic share 5.3%
- #8 Z.ai: GLM 5.2 (free) - Agentic traffic share 3.3%
- #9 Google: Gemini 3.8 Flash (batch) - Agentic traffic share 2.2%
- #10 OpenAI: GPT-6 Astra (batch) - Agentic traffic share 1.6%
| # | Model | Agentic traffic share |
|---|---|---|
| 01 | Z.ai: GLM 5.3 Flash (batch) | 19.3% |
| 02 | DeepSeek: DeepSeek V4 Flash 0423 | 18.0% |
| 03 | DeepSeek: DeepSeek V4.1 Flash (batch) | 16.4% |
| 04 | OpenAI: GPT-5.6 Luna (batch) | 14.6% |
| 05 | Tencent: Hy4 preview | 10.8% |
| 06 | Tencent: Hy3 | 5.3% |
| 07 | Xiaomi: MiMo-V2.5 | 5.3% |
| 08 | Z.ai: GLM 5.2 (free) | 3.3% |
| 09 | Google: Gemini 3.8 Flash (batch) | 2.2% |
| 10 | OpenAI: GPT-6 Astra (batch) | 1.6% |
| 11 | Meta: Muse Spark 1.3 Contributor | 0.9% |
| 12 | Google: Gemma 4 26B A4B (free) | 0.6% |
| 13 | Google: Gemini 2.5 Flash Lite (batch) | 0.5% |
| 14 | OpenAI: GPT-4o-mini (batch) | 0.3% |
| 15 | Google: Gemini 3.5 Flash Lite (batch) | 0.3% |
| 16 | OpenAI: GPT-5.6 Terra (batch) | 0.2% |
| 17 | Perplexity: Sonar | 0.1% |
| 18 | MiniMax: MiniMax M3 | 0.1% |
Source: OpenRouter (openrouter.ai/rankings), as of September 23, 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.
