Most-used AI models for debugging
DeepSeek V4 Flash 0423 is the most-used model for debugging on OpenRouter, handling 11.1% of requests classified as debugging over the last 7 days.
This ranks models by how often they are actually chosen for debugging 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 | 11.1% | 8.9% | $0.140 | 1.0M |
| 2 |  GLM 5.2 | 9.7% | 7.2% | $0.760 | 1.0M |
| 3 |  MiMo-V2.5 | 9.5% | 13.2% | $0.140 | 1.1M |
| 4 | DeepSeek V4 Pro | 8.9% | 7.0% | $0.435 | 1.0M |
| 5 |  Claude Sonnet 4.6 | 5.5% | 1.9% | $3.00 | 1M |
| 6 | DeepSeek V4 Flash 0423 | 4.8% | 7.9% | $0.140 | 1.0M |
| 7 |  GPT-5.6 Luna | 3.6% | 6.5% | $0.100 | 1.1M |
| 8 |  Step 3.7 Flash | 3.2% | 4.8% | $0.200 | 262K |
| 9 | nemotron-3-ultra-550b-a55b-20260604:free | 3.2% | 6.2% | — | — |
| 10 | ling-3.0-flash-20260723:free | 3.1% | 3.8% | — | — |
Reading the two columns together: DeepSeek V4 Flash 0423, GPT-5.6 Luna, nemotron-3-ultra-550b-a55b-20260604:free take a noticeably larger share of tokens than of requests — meaning they are being used for the longer, heavier debugging jobs rather than quick one-shot calls.
Frequently asked
DeepSeek V4 Flash 0423 is the most-used model for debugging on OpenRouter, handling 11.1% of requests classified as debugging over the last 7 days. It is followed by GLM 5.2 (9.7%) and MiMo-V2.5 (9.5%).
No. This ranks by how often each model is actually chosen for debugging 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.
Debugging accounts for 1.9% of classified requests and 6.8% 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.