DeepSeek R1 Distilled Qwen 14B
Qwen 14B distilled with reasoning capabilities from Deepseek R1. Outperforms GPT-4o in math & matches o1-mini on coding.
About model
DeepSeek R1 Distilled Qwen 14B excels at complex problem-solving and reasoning tasks. It leverages large-scale reinforcement learning to develop chain-of-thought capabilities. Suitable for researchers and developers seeking advanced reasoning models.
Model | AIME 2025 | GPQA Diamond | HLE | LiveCodeBench | MATH500 | SWE-bench verified |
|---|---|---|---|---|---|---|
DeepSeek R1 Distilled Qwen 14B | 65.0% | 78.3% | Related open-source models | Competitor closed-source models | ||
90.5% | 34.2% | 78.7% | ||||
83.3% | 24.9% | 99.2% | 62.3% | |||
76.8% | 96.4% | 48.9% | ||||
49.2% | 2.7% | 32.3% | 89.3% | 31.0% |
API usage
Endpoint:
- TypeChatReasoning
- Main use casesChatReasoning
- FeaturesJSON Mode
- Fine tuningSupported
- DeploymentServerlessOn-Demand DedicatedMonthly Reserved
- Parameters14.8B
- Input price
$0.18 / 1M tokens
- Output price
$0.18 / 1M tokens
- Input modalitiesText
- Output modalitiesText
- ReleasedJanuary 20, 2025
- Last updatedSeptember 9, 2025
- Quantization levelFP16
- External link
- CategoryChat