LiveCodeBench v5
We show this table for reference; we do not rank on it.
LiveCodeBench v5 is a named release and date-window slice. BenchLM keeps explicitly labeled v5 rows outside the rolling weighted lane.
Benchmark score on LiveCodeBench v5 — November 13, 2026
We compile the LiveCodeBench v5 rows from provider self-reports. Ling 3.0 Flash leads the table at 82.8%, followed by Qwen3 235B-A22B (Reasoning) (70.7%) and Qwen3 32B (Reasoning) (65.7%). We do not use these results to rank models overall.
Ling 3.0 Flash
InclusionAI
Reported by InclusionAIQwen3 235B-A22B (Reasoning)
Alibaba
Reported by AlibabaQwen3 32B (Reasoning)
Alibaba
Reported by Alibaba9 modelsCodingCurrentDisplay onlyUpdated November 13, 2026
Benchmark score table (9 models)
All 9 results on this page were reported by the labs that built the models.
| Rank | Model / configuration | Score | Who ran it | Parameters (B) | Open / closed |
|---|---|---|---|---|---|
| 1 | Ling 3.0 FlashInclusionAIReported by InclusionAI | 82.8% | Reported by InclusionAI | Not reported | Open |
| 2 | Qwen3 235B-A22B (Reasoning)AlibabaReported by Alibaba | 70.7% | Reported by Alibaba | Not reported | Open |
| 3 | Qwen3 32B (Reasoning)AlibabaReported by Alibaba | 65.7% | Reported by Alibaba | Not reported | Open |
| 4 | Nemotron 3 Nano Omni 30B A3BNVIDIAReported by NVIDIA | 63.2% | Reported by NVIDIA | Not reported | Open |
| 5 | QwQ-32BAlibabaReported by Alibaba | 62.7% | Reported by Alibaba | Not reported | Open |
| 6 | Magistral MediumMistralReported by Mistral | 59.4% | Reported by Mistral | Not reported | Closed |
| 7 | Magistral SmallMistralReported by Mistral | 55.8% | Reported by Mistral | Not reported | Open |
| 8 | Qwen3 235B-A22BAlibabaReported by Alibaba | 35.3% | Reported by Alibaba | Not reported | Open |
| 9 | Qwen3 32BAlibabaReported by Alibaba | 31.3% | Reported by Alibaba | Not reported | Open |
- Lab-reported
- Benchmark-run
- Independent
How to read this leaderboard
Editorial review by Glevd · 2026-07-15
Use this route as a version-specific source receipt. Match the date window, scenario, pass@k or average-at-k metric, sampling settings, and execution policy before comparing another v5 result.
Operator receipt: 9 sourced rows are currently displayable on this page; the leading published row is Ling 3.0 Flash at 82.8%.
Honest limit: One sourced row cannot establish a leader. The provider report supplies the value, but BenchLM did not rerun the evaluation. LiveCodeBench also does not measure work inside an existing repository.
Among the reported LiveCodeBench v5 rows, Ling 3.0 Flash is first at 82.8%. The third row is 17.1 points behind. The broader top-10 range is 51.5 points, so the table still separates the published systems.
9 models have been evaluated on LiveCodeBench v5. The benchmark falls in the Coding category. LiveCodeBench v5 is currently displayed for reference but excluded from the scoring formula, so it does not directly affect overall rankings.
About LiveCodeBench v5
Year
2025
Tasks
July 2024 to May 2025 release window
Format
Provider-published v5 competitive programming result
Difficulty
Competitive programming level
The current sourced row identifies the July 2024 through May 2025 v5 window. It is preserved as published instead of being treated as interchangeable with v6 or an unspecified rolling snapshot.
Freshness and provenance
Version
LiveCodeBench v5 2025
Refresh cadence
Quarterly
Staleness state
Current
Question availability
Public benchmark set
BenchLM uses freshness metadata to decide whether a benchmark should still be treated as a strong differentiator, a benchmark to watch, or a display-only reference. For the full scoring policy, see the BenchLM methodology page.
Questions
What does LiveCodeBench v5 measure?
LiveCodeBench v5 is a named release and date-window slice. BenchLM keeps explicitly labeled v5 rows outside the rolling weighted lane.
Which model scores highest on LiveCodeBench v5?
Ling 3.0 Flash by InclusionAI currently leads with a score of 82.8% on LiveCodeBench v5.
How many models are evaluated on LiveCodeBench v5?
9 AI models have published results on LiveCodeBench v5 in the BenchLM catalog.
Know when it’s worth switching models
The model to choose, the cheaper alternative, and the release we would wait on.
Read a sample issueJoin 7,000+ subscribers.
One email each week. Unsubscribe anytime.