Most-used AI models for roleplay & fiction
DeepSeek V4 Flash 0423 is the most-used model for roleplay & fiction on OpenRouter, handling 42.2% of requests classified as roleplay & fiction over the last 7 days.
This ranks models by how often they are actually chosen for roleplay & fiction 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 | 42.2% | 31.2% | $0.140 | 1.0M |
| 2 | DeepSeek V4 Pro | 5.2% | 7.3% | $0.435 | 1.0M |
| 3 |  Gemini 2.5 Flash Lite | 4.2% | 3.0% | $0.100 | 1.0M |
| 4 | DeepSeek V3.2 | 4.1% | 6.0% | $0.269 | 164K |
| 5 | Gemma 4 31B | 3.4% | 7.1% | $0.100 | 262K |
| 6 | ILing-2.6-flash | 3.4% | 1.6% | $0.010 | 262K |
| 7 |  Llama 4 Maverick | 3.2% | 0.50% | $0.200 | 1.0M |
| 8 | Gemma 4 26B A4B | 3.1% | 4.0% | $0.070 | 262K |
| 9 | DeepSeek V3.1 Terminus | 2.9% | 0.70% | $0.270 | 164K |
| 10 |  Mistral Nemo | 2.6% | 1.1% | $0.019 | 131K |
Reading the two columns together: Gemma 4 31B takes a noticeably larger share of tokens than of requests — meaning it is being used for the longer, heavier roleplay & fiction jobs rather than quick one-shot calls.
Frequently asked
DeepSeek V4 Flash 0423 is the most-used model for roleplay & fiction on OpenRouter, handling 42.2% of requests classified as roleplay & fiction over the last 7 days. It is followed by DeepSeek V4 Pro (5.2%) and Gemini 2.5 Flash Lite (4.2%).
No. This ranks by how often each model is actually chosen for roleplay & fiction 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.
Roleplay & Fiction accounts for 12.4% of classified requests and 6.1% 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.