Qwen: Qwen3 14B

Text input Text output
Author's Description

Qwen3-14B is a dense 14.8B parameter causal language model from the Qwen3 series, designed for both complex reasoning and efficient dialogue. It supports seamless switching between a "thinking" mode for tasks like math, programming, and logical inference, and a "non-thinking" mode for general-purpose conversation. The model is fine-tuned for instruction-following, agent tool use, creative writing, and multilingual tasks across 100+ languages and dialects. It natively handles 32K token contexts and can extend to 131K tokens using YaRN-based scaling.

Key Specifications
Cost
$$$
Context
40K
Parameters
14B
Released
Apr 28, 2025
Speed
Ability
Reliability
Supported Parameters

This model supports the following parameters:

Stop Presence Penalty Temperature Seed Response Format Frequency Penalty Max Tokens Include Reasoning Tool Choice Top P Min P Tools Reasoning
Features

This model supports the following features:

Tools Reasoning Response Format
Performance Summary

Qwen3 14B demonstrates a balanced performance profile, excelling in reliability and offering competitive pricing, though its speed is moderate. It ranks in the 28th percentile for speed, indicating it is not among the fastest models, with durations for benchmarks like Instruction Following and Coding falling into the 16th percentile. Conversely, its pricing is competitive, placing it in the 59th percentile. A standout feature is its exceptional reliability, achieving a perfect 100th percentile, meaning it consistently provides usable responses with minimal technical failures. In terms of specific benchmarks, Qwen3 14B shows strong capabilities in Reasoning, achieving 96.0% accuracy and ranking in the 91st percentile, notably being the most accurate model at its price point. It also performs well in Coding (91.0% accuracy, 84th percentile) and General Knowledge (98.8% accuracy, 70th percentile). While its Ethics performance is solid at 99.0% accuracy (61st percentile), its Instruction Following accuracy is moderate at 50.5% (51st percentile), suggesting an area for potential improvement in handling complex, multi-layered instructions. Email Classification is also consistent at 97.0% accuracy (51st percentile). Overall, Qwen3 14B is a highly reliable and cost-effective model, particularly strong in complex reasoning and coding tasks, making it suitable for applications requiring high accuracy and dependable operation.

Model Pricing

Current Pricing

Feature Price (per 1M tokens)
Prompt $0.06
Completion $0.24

Price History

Available Endpoints
Provider Endpoint Name Context Length Pricing (Input) Pricing (Output)
DeepInfra
DeepInfra | qwen/qwen3-14b-04-28 40K $0.06 / 1M tokens $0.24 / 1M tokens
Parasail
Parasail | qwen/qwen3-14b-04-28 40K $0.06 / 1M tokens $0.24 / 1M tokens
Nebius
Nebius | qwen/qwen3-14b-04-28 40K $0.08 / 1M tokens $0.24 / 1M tokens
Parasail
Parasail | qwen/qwen3-14b-04-28 40K $0.06 / 1M tokens $0.25 / 1M tokens
NextBit
NextBit | qwen/qwen3-14b-04-28 40K $0.06 / 1M tokens $0.24 / 1M tokens
Benchmark Results
Benchmark Category Reasoning Free Executions Accuracy Cost Duration
Other Models by qwen