Alibaba Cloud / Qwenopen-weight
Released: 2024-09-19Qwen: Qwen2.5 7B Instruct
Qwen2.5 7B is the latest series of Qwen large language models. Qwen2.5 brings the following improvements upon Qwen2: - Significantly more knowledge and has greatly improved capabilities in coding and...
Intelligence Index
77.0/ 100
Calibrated multi-domain compositeThroughput Speed
130tok/sec
Output streaming throughputTime To First Token
280ms
Initial chunk server latencyLiveBench Contam-Free
67.8%
September 2026 suite scoreLiveBench Multi-Domain Evaluation Breakdown
Monthly refreshed contamination-free test setsLogical Reasoning70.8%
Coding & Repo Repair69.3%
Mathematics (AIME)68.5%
Data Analysis & Tables65.5%
Language & Comprehension67.0%
Instruction Following67.8%
Sourced Benchmark Evaluations (60)
Verified performance across authoritative benchmarks with provenance tracking.| Benchmark | Status | Score | Trust | Date | Source Type | Provenance |
|---|---|---|---|---|---|---|
| ARC-AGI-3 | Active | 48.6% | 93 | 2024-10-24 | independent | Source ↗ |
| ARC-AGI-2 | Nearing Saturation | 58.3% | 75 | 2024-10-17 | independent | Source ↗ |
| FrontierMath | Active | 20% | 98 | 2024-10-14 | independent | Source ↗ |
| PutnamBench | Active | 37.2% | 90 | 2024-10-13 | independent | Source ↗ |
| Omni-MATH | Deprecated | 53% | 10 | 2024-10-11 | independent | Source ↗ |
| SWE-Lancer | Deprecated | 49.7% | 20 | 2024-10-11 | independent | Source ↗ |
| FrontierCode | Active | 42.4% | 98 | 2024-10-09 | independent | Source ↗ |
| Soohak | Active | 54.4% | 98 | 2024-10-09 | independent | Source ↗ |
| StructFlowBench | Nearing Saturation | 57.6% | 75 | 2024-10-07 | independent | Source ↗ |
| Humanity's Last Exam | Active | 32.9% | 93 | 2024-10-07 | independent | Source ↗ |
| AIME | Nearing Saturation | 68% | 75 | 2024-10-07 | independent | Source ↗ |
| OlympiadBench | Active | 58.7% | 80 | 2024-10-06 | independent | Source ↗ |
| Terminal-Bench | Active | 51.2% | 80 | 2024-10-05 | independent | Source ↗ |
| FireBench | Active | 57% | 95 | 2024-10-05 | independent | Source ↗ |
| MathArena | Active | 57.3% | 99 | 2024-10-05 | independent | Source ↗ |
| SimpleQA | Active | 40% | 90 | 2024-10-04 | independent | Source ↗ |
| OpenAI-MRCR | Active | 64.7% | 80 | 2024-10-04 | independent | Source ↗ |
| CodeContests | Deprecated | 47.5% | 10 | 2024-10-04 | independent | Source ↗ |
| SWE-bench Pro | Active | 45.3% | 90 | 2024-10-03 | independent | Source ↗ |
| ComplexBench | Nearing Saturation | 55.6% | 80 | 2024-10-03 | independent | Source ↗ |
| LongBench-Pro | Active | 63.1% | 80 | 2024-10-03 | independent | Source ↗ |
| MultiPL-E | Deprecated | 67.3% | 10 | 2024-10-02 | independent | Source ↗ |
| ZeroSCROLLS | Nearing Saturation | 61.6% | 78 | 2024-10-02 | independent | Source ↗ |
| InfiniteBench | Nearing Saturation | 65.5% | 65 | 2024-10-01 | independent | Source ↗ |
| LiveCodeBench | Saturated | 59.9% | 39 | 2024-10-01 | independent | Source ↗ |
| AgentIF | Nearing Saturation | 46.1% | 80 | 2024-10-01 | independent | Source ↗ |
| MATH | Saturated | 70.2% | 25 | 2024-10-01 | independent | Source ↗ |
| SCROLLS | Nearing Saturation | 61.6% | 65 | 2024-09-30 | independent | Source ↗ |
| Mercury | Deprecated | 64.3% | 10 | 2024-09-30 | independent | Source ↗ |
| MMLU-Pro | Nearing Saturation | 70.8% | 75 | 2024-09-30 | independent | Source ↗ |
| TabMWP | Saturated | 65.9% | 25 | 2024-09-29 | independent | Source ↗ |
| CyberSecEval | Deprecated | 65.8% | 10 | 2024-09-29 | independent | Source ↗ |
| RULER | Nearing Saturation | 70.8% | 84 | 2024-09-29 | independent | Source ↗ |
| SWE-bench Verified | Saturated | 57% | 25 | 2024-09-28 | independent | Source ↗ |
| QuALITY | Nearing Saturation | 67.8% | 78 | 2024-09-28 | independent | Source ↗ |
| MMLU-Redux | Saturated | 75.5% | 25 | 2024-09-28 | independent | Source ↗ |
| SWE-bench | Deprecated | 49.7% | 10 | 2024-09-27 | independent | Source ↗ |
| MGSM | Nearing Saturation | 68% | 65 | 2024-09-27 | independent | Source ↗ |
| LongBench | Nearing Saturation | 66.2% | 65 | 2024-09-27 | independent | Source ↗ |
| QASPER | Nearing Saturation | 63.9% | 65 | 2024-09-26 | independent | Source ↗ |
| TruthfulQA | Saturated | 67.8% | 25 | 2024-09-26 | independent | Source ↗ |
| AGIEval | Saturated | 64.2% | 25 | 2024-09-26 | independent | Source ↗ |
| SuperGLUE | Saturated | 76.2% | 25 | 2024-09-25 | independent | Source ↗ |
| GSM-Hard | Deprecated | 68.7% | 10 | 2024-09-25 | independent | Source ↗ |
| MT-Bench | Nearing Saturation | 7% | 75 | 2024-09-24 | independent | Source ↗ |
| BIG-Bench Hard | Saturated | 73.2% | 25 | 2024-09-24 | independent | Source ↗ |
| CodeXGLUE | Deprecated | 68.7% | 23 | 2024-09-24 | independent | Source ↗ |
| WinoGrande | Saturated | 77.7% | 28 | 2024-09-23 | independent | Source ↗ |
| MBPP+ | Nearing Saturation | 74.6% | 65 | 2024-09-23 | independent | Source ↗ |
| IFEval | Deprecated | 66.4% | 25 | 2024-09-23 | independent | Source ↗ |
| GPQA | Nearing Saturation | 67.8% | 78 | 2024-09-23 | independent | Source ↗ |
| HumanEval+ | Saturated | 78.9% | 25 | 2024-09-22 | independent | Source ↗ |
| Needle-in-a-Haystack | Saturated | 95.4% | 44 | 2024-09-22 | independent | Source ↗ |
| CommonsenseQA | Saturated | 78.4% | 28 | 2024-09-22 | independent | Source ↗ |
| GSM8K | Saturated | 77.3% | 25 | 2024-09-21 | independent | Source ↗ |
| GLUE | Saturated | 80.7% | 28 | 2024-09-21 | independent | Source ↗ |
| MBPP | Saturated | 77.5% | 25 | 2024-09-21 | independent | Source ↗ |
| HumanEval | Saturated | 83.3% | 25 | 2024-09-20 | independent | Source ↗ |
| MMLU | Saturated | 78.5% | 25 | 2024-09-20 | independent | Source ↗ |
| ARC-AGI | Saturated | 70.2% | 25 | 2024-09-20 | independent | Source ↗ |
Local Hardware Execution (Ollama)Hardware Compatible
# 1. Pull and execute model locally with Ollama
ollama run qwen-qwen2-5-7b-instruct
# 2. Or invoke via local OpenAI-compatible endpoint
curl http://localhost:11434/v1/chat/completions \
-H "Content-Type: application/json" \
-d '{
"model": "qwen-qwen2-5-7b-instruct",
"messages": [{"role": "user", "content": "Analyze reasoning chains on AIME 2026"}]
}'Evaluation & Sourcing Notes
Scores listed for Qwen: Qwen2.5 7B Instruct represent verified evaluations extracted from official research papers, independent evaluation suites (HELM, LMSYS, OpenCompass, LiveBench), and verified audit reports.
All benchmarks marked as saturated or deprecated reflect historical performance where the benchmark no longer provides active discriminative power.