SWE Bench Multilingual
AI benchmark results and model performance
Type: percentageCode
Recorded results
35
Average score
67.61%
Score range
39.33% - 89.50%
Leading model
89.50% - Claude Opus 5
Scores over time
Individual benchmark scores plotted by date.
Model results
| Organisation | Models | Reported | Top score | Info | Self reported | Source |
|---|---|---|---|---|---|---|
| Claude Opus 5 | 24 Jul 2026 | 89.50% | 300 problems across 9 programming languages; adaptive thinking at max effort; average of 5 trials | Yes | Source | |
| Claude Mythos Preview | 07 Apr 2026 | 87.30% | - | Yes | Source | |
| Claude Opus 4.7 | 16 Apr 2026 | 80.50% | - | Yes | Source | |
| Laguna S 2.1 | 21 Jul 2026 | 78.50% | mean pass@1 | Yes | Source | |
| Qwen 3.7 Max | 21 May 2026 | 78.30% | Internal agent scaffold | Yes | Source | |
| Claude Sonnet 5 | 30 Jun 2026 | 78.30% | SWE-bench Multilingual; standard configuration; average over five trials | Yes | Source | |
| Beam | 05 Oct 2026 | 78% | Reflection-reported launch score. Source row: SWEBench Multilingual. Accessed 2026-10-05T22:51:12.456Z. Harness, reasoning effort, sampling and trial count are not specified per row. Not independently reproduced. | Yes | Source | |
| Claude Opus 4.6 | 05 Feb 2026 | 77.83% | - | Yes | Source | |
| MiniMax M2.7 | 18 Mar 2026 | 76.50% | - | Yes | Source | |
| Claude Opus 4.5 | 24 Nov 2025 | 76.20% | Avg@5 | Yes | Source | |
| Claude Sonnet 4.6 | 17 Feb 2026 | 75.90% | - | Yes | Source | |
| Qwen 3.6 Plus | 01 Apr 2026 | 73.80% | - | Yes | Source | |
| Kimi K2.5 | 27 Jan 2026 | 73% | - | Yes | Source | |
| MiniMax M2.1 | 23 Dec 2025 | 72.50% | Outperforms Claude Sonnet 4.5 (68%) and DeepSeek V3.2 (70.2%) | Yes | Source | |
| Seed 2.0 Pro | 14 Feb 2026 | 71.70% | Seed2 official benchmark table | SWE Multilingual | Yes | Source | |
| MiMo V2 Flash | 16 Dec 2025 | 71.70% | - | Yes | Source | |
| MiMo V2 Pro | 18 Mar 2026 | 71.70% | - | Yes | Source | |
| MiMo V2 TTS (Free) | 18 Mar 2026 | 71.70% | inferred modality/version alias from mimo-v2-pro | Yes | Source | |
| Qwen 3.5 397B A17B | 16 Feb 2026 | 69.30% | - | Yes | Source | |
| GLM 4.7 | 22 Dec 2025 | 66.70% | - | Yes | Source | |
| Seed 2.0 Lite | 14 Feb 2026 | 64.40% | Seed2 official benchmark table | SWE Multilingual | Yes | Source | |
| Laguna XS 2.1 | 02 Jul 2026 | 63.10% | mean pass@1 | Yes | Source | |
| Kimi K2 Thinking | 06 Nov 2025 | 61.10% | inferred alias from kimi-k2-thinking-0905 | Yes | Source | |
| DeepSeek OCR 2 | - | 57.90% | inferred family alias from deepseek-v3.2-exp (score=0.3809; benches=14) | Yes | Source | |
| DeepSeek V3.2 Exp | 29 Sept 2025 | 57.90% | - | Yes | Source |
