# LiveCodeBench v6

> LiveCodeBench v6 is a named release slice used in provider comparison tables. Keeping it separate prevents v6 results from being mixed into older or rolling LiveCodeBench windows.

Canonical page: https://benchlm.ai/benchmarks/livecodebench-v6

- Category: [Coding](/coding)
- Last updated: September 10, 2026

## About LiveCodeBench v6

- Year: 2026
- Tasks: Fresh programming problems
- Format: Provider-published v6 competitive programming results
- Difficulty: Competitive programming level
- Paper: [LiveCodeBench official repository and release documentation](https://github.com/LiveCodeBench/LiveCodeBench)

Providers often publish a specific release or date window instead of the rolling aggregate. This route contains rows explicitly labeled v6 by their sources and excludes them from the weighted legacy lane.

LiveCodeBench v6 is currently displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Leaderboard (29 models)

| Rank | Model | Creator | Score |
|------|-------|---------|-------|
| 1 | [Sakana Fugu-Ultra](/models/sakana-fugu-ultra) | Sakana AI | 93.2% |
| 2 | [Sakana Fugu](/models/sakana-fugu) | Sakana AI | 92.9% |
| 3 | [Solar Open 2](/models/solar-open-2) | Upstage | 92.4% |
| 4 | [Qwen3.8-Flash-Next](/models/qwen3-8-flash-next) | Alibaba | 91.9% |
| 5 | [dots3-note Preview](/models/dots3-note-preview) | Dots Studio | 91.5% |
| 6 | [Qwen3.8-27B](/models/qwen3-8-27b) | Alibaba | 90.3% |
| 7 | [Kimi K2.6](/models/kimi-2-6) | Moonshot AI | 89.6% |
| 8 | [Nemotron 3 Ultra](/models/nemotron-3-ultra) | NVIDIA | 89.0% |
| 9 | [BTL-3](/models/btl-3) | Bad Theory Labs | 88.1% |
| 10 | [MAI-Thinking-1](/models/mai-thinking-1) | Microsoft | 87.7% |
| 11 | [Qwen3.6 Plus](/models/qwen3-6-plus) | Alibaba | 87.1% |
| 12 | [Kimi K2.5](/models/kimi-k2-5) | Moonshot AI | 85.0% |
| 13 | [Claude Opus 4.5](/models/claude-opus-4-5) | Anthropic | 84.8% |
| 14 | [A.X K2](/models/a-x-k2) | SK Telecom | 84.0% |
| 15 | [Qwen3.5 397B](/models/qwen3-5-397b) | Alibaba | 83.6% |
| 16 | [Granite 4.2 30B](/models/granite-4-2-30b) | IBM | 75.8% |
| 17 | [Granite 4.2 8B](/models/granite-4-2-8b) | IBM | 73.2% |
| 18 | [Gemma 4 12B](/models/gemma-4-12b) | Google | 72.0% |
| 19 | [Mellum2-12B-A2.5B-Thinking](/models/mellum2-12b-a2-5b-thinking) | JetBrains | 69.9% |
| 20 | [Granite 4.2 3B](/models/granite-4-2-3b) | IBM | 69.7% |
| 21 | [MiniCPM5-2B](/models/minicpm5-2b) | OpenBMB | 69.1% |
| 22 | [BTL-4](/models/btl-4) | Bad Theory Labs | 66.1% |
| 23 | [ZAYA1-8B](/models/zaya1-8b) | Zyphra | 65.8% |
| 24 | [ZAYA1-74B-Preview](/models/zaya1-74b-preview) | Zyphra | 65.7% |
| 25 | [Agents-A1-4B](/models/agents-a1-4b) | InternScience | 59.6% |
| 26 | [LFM2.5-2.6B](/models/lfm2-5-2-6b) | LiquidAI | 59.4% |
| 27 | [Mellum2-12B-A2.5B-Instruct](/models/mellum2-12b-a2-5b-instruct) | JetBrains | 37.2% |
| 28 | [MiniCPM5-1B](/models/minicpm5-1b) | OpenBMB | 33.5% |
| 29 | [LLaDA2.2-mini](/models/llada2-2-mini) | InclusionAI | 28.1% |

## FAQ

### What does LiveCodeBench v6 measure?

LiveCodeBench v6 is a named release slice used in provider comparison tables. Keeping it separate prevents v6 results from being mixed into older or rolling LiveCodeBench windows.

### Which model scores highest on LiveCodeBench v6?

Sakana Fugu-Ultra by Sakana AI currently leads with a score of 93.2% on LiveCodeBench v6.

### How many models are evaluated on LiveCodeBench v6?

29 AI models have been evaluated on LiveCodeBench v6 on BenchLM.

### Does LiveCodeBench v6 affect BenchLM's overall score?

Not directly. LiveCodeBench v6 is still displayed on BenchLM for reference, but it is excluded from the weighted scoring formula.

## Compare Top Models on LiveCodeBench v6

- [Sakana Fugu-Ultra vs Sakana Fugu](/compare/sakana-fugu-vs-sakana-fugu-ultra)
- [Sakana Fugu vs Solar Open 2](/compare/sakana-fugu-vs-solar-open-2)
- [Solar Open 2 vs Qwen3.8-Flash-Next](/compare/qwen3-8-flash-next-vs-solar-open-2)
- [Qwen3.8-Flash-Next vs dots3-note Preview](/compare/dots3-note-preview-vs-qwen3-8-flash-next)