Most-used AI models for multi-step planning
DeepSeek V4 Flash 0423 is the most-used model for multi-step planning on OpenRouter, handling 17.1% of requests classified as multi-step planning over the last 7 days.
This ranks models by how often they are actually chosen for multi-step planning in production traffic, not by benchmark score. Adoption reflects price, latency and availability as much as raw capability — which is exactly why it often disagrees with the benchmark leaderboards, and why it is worth reading alongside them.
| # | Model | Request share | Token share | Input /M | Context |
|---|---|---|---|---|---|
| 1 |  DeepSeek V4 Flash 0423 | 17.1% | 11.7% | $0.140 | 1.0M |
| 2 |  Hy3 | 5.2% | 8.6% | $0.132 | 262K |
| 3 | DeepSeek V4 Pro | 5.0% | 5.4% | $0.435 | 1.0M |
| 4 |  GLM 5.2 | 4.5% | 7.1% | $0.760 | 1.0M |
| 5 | DeepSeek V4 Flash 0423 | 4.5% | 8.4% | $0.140 | 1.0M |
| 6 |  GPT-5.6 Luna | 3.9% | 7.6% | $0.100 | 1.1M |
| 7 |  MiMo-V2.5 | 3.6% | 6.7% | $0.140 | 1.1M |
| 8 |  Gemini 2.5 Flash Lite | 3.5% | 0.40% | $0.100 | 1.0M |
| 9 |  Kimi K3 | 3.0% | 4.9% | $3.00 | 1.0M |
| 10 | Gemini 3 Flash Preview | 3.0% | 0.60% | $0.500 | 1.0M |
Reading the two columns together: Hy3, GLM 5.2, DeepSeek V4 Flash 0423 take a noticeably larger share of tokens than of requests — meaning they are being used for the longer, heavier multi-step planning jobs rather than quick one-shot calls.
Frequently asked
DeepSeek V4 Flash 0423 is the most-used model for multi-step planning on OpenRouter, handling 17.1% of requests classified as multi-step planning over the last 7 days. It is followed by Hy3 (5.2%) and DeepSeek V4 Pro (5.0%).
No. This ranks by how often each model is actually chosen for multi-step planning in production traffic through OpenRouter — real-world adoption, which reflects price and availability as much as capability. For capability-based rankings, see the benchmark leaderboards.
Multi-step Planning accounts for 2.5% of classified requests and 6.2% of classified tokens on OpenRouter over the trailing 7 days.
Other tasks
Source: OpenRouter (openrouter.ai/rankings), as of 2026-08-04. Shares are of classified, sampled traffic over a trailing 7-day window; absolute volumes are not published.