Compare AI models
| Specifications | ||
ParametersThe learned values that shape how a model responds. B means billion and T means trillion. For mixture-of-experts models, the total count is larger than the number active for each token. Parameter count alone does not measure quality. | 0.5B/1.5B/3B/7B/14B/32B/72B | — |
Context windowHow much information the model can work with in one request, measured in tokens — pieces of words or other data. This includes your question, attachments, conversation and reply. Limits can vary by model variant and provider. | 128k | 1M |
| API pricingUSD per 1M tokens · lower wins | ||
Cheapest inputLowest input rate across every provider on OpenRouter, the lab included. The cheapest endpoint may run a quantised build or a shorter context — see "Available from" on the model page. | $0.288DeepInfra | — |
Cheapest outputLowest output rate across every provider on OpenRouter, the lab included. May come from a different provider than the cheapest input. | $0.32DeepInfra | — |
| $92.00DeepInfra · fp8 | — | |
| Benchmarks | ||
LiveCodeBenchCompetitive coding — Coding problems published so recently the AI can't have seen them in training — a contamination-free test of raw programming skill. Higher is better. | 55.5% | 37.8% |
HumanEvalFunction synthesis — 164 small Python problems: the AI is given a function's description and has to write the working function. This was the coding benchmark of the GPT-3.5 and GPT-4 era, before the field moved to fixing real bugs in real repositories. Higher is better. | 86.6% | 86.6% |
IFEvalAgentic tasks — following verifiable instructions in generated responses | 84.1% | 76.3% |
MATHReasoning — accuracy on the original competition mathematics test set | 83.1% | 81.1% |
MMLU-ProKnowledge — accuracy on the ten-option MMLU-Pro benchmark | 71.1% | 64.5% |
GSM8KGrade-school math — Grade-school maths word problems that take a few steps of arithmetic to work through. It separated the models of 2022 and 2023 sharply, then saturated. One caveat on the historical numbers: OpenAI included part of the GSM8K training set in GPT-4's pre-training mix, so GPT-4's score is not a clean few-shot result. Higher is better. | 95.8% | 93.8% |
| Overview | ||
| Company | Qwen | Qwen |
| Release date | Sep 19 2024 | Nov 15 2024 |
| Access | Open Weight | Closed |
| Model details | View model | View model |
Other comparisons
Qwen2.5vsClaude Fable 5.2Qwen2.5-Turbo (1M)vsClaude Fable 5.2Qwen2.5vsGPT-6.2 AstraQwen2.5-Turbo (1M)vsGPT-6.2 AstraQwen2.5vsGemini 4 NeonQwen2.5-Turbo (1M)vsGemini 4 NeonQwen2.5vsMuse EmberQwen2.5-Turbo (1M)vsMuse EmberQwen2.5vsGrok 4.7Qwen2.5-Turbo (1M)vsGrok 4.7Qwen2.5vsDeepSeek-V4.1-Flash-1017Qwen2.5-Turbo (1M)vsDeepSeek-V4.1-Flash-1017Frequently asked questions
Qwen2.5 leads Qwen2.5-Turbo (1M) on 5 of the 6 benchmarks they both report. Qwen2.5 shipped 57 days before Qwen2.5-Turbo (1M), so benchmark comparisons should account for the intervening progress.
Context windows are 128k (Qwen2.5) vs 1M (Qwen2.5-Turbo (1M)). Qwen2.5 is open weight, while Qwen2.5-Turbo (1M) is closed.
On LiveCodeBench, Qwen2.5 leads at 55.5% vs Qwen2.5-Turbo (1M) at 37.8%. On HumanEval, both models score 86.6%. On IFEval, Qwen2.5 leads at 84.1% vs Qwen2.5-Turbo (1M) at 76.3%. On MATH, Qwen2.5 leads at 83.1% vs Qwen2.5-Turbo (1M) at 81.1%. On MMLU-Pro, Qwen2.5 leads at 71.1% vs Qwen2.5-Turbo (1M) at 64.5%. On GSM8K, Qwen2.5 leads at 95.8% vs Qwen2.5-Turbo (1M) at 93.8%.