A multilingual software-engineering benchmark covering realistic repository tasks across nine programming languages.
Browse the latest scores, model modes, release dates, and parameter sizes for SWE-bench Multilingual.
Data sourced primarily from official releases (GitHub, Hugging Face, papers), then benchmark leaderboards, then third-party evaluators. Learn about our data methodology
| Rank | Model | License | |||
|---|---|---|---|---|---|
![]() Claude Mythos Preview Extended ThinkingTools | 87.30 | 2026-04-07 | Unknown | Closed | |
![]() Composer 2.5 Thinking Enabled | 79.80 | 2026-05-18 | Unknown | Closed | |
![]() Qwen3.7-Max-Preview Thinking EnabledTools | 78.30 | 2026-05-20 | Unknown | Closed | |
4 | ![]() Kimi K2.6 Thinking EnabledTools | 76.70 | 2026-04-20 | 1000B | Free Commercial |
5 | ![]() DeepSeek-V4-Pro Thinking Level · Extra HighTools | 76.20 | 2026-04-24 | 1600B | Free Commercial |
6 | ![]() Hy3 Thinking Level · HighTools | 75.80 | 2026-07-06 | 295B | Free Commercial |
7 | ![]() DeepSeek-V4-Pro Thinking Level · HighTools | 74.10 | 2026-04-24 | 1600B | Free Commercial |
8 | ![]() Qwen 3.6 Plus Preview Thinking Enabled | 73.80 | 2026-03-31 | Unknown | Closed |
9 | ![]() Qwen3.6-Max-Preview Thinking EnabledTools | 73.80 | 2026-04-18 | Unknown | Closed |
10 | ![]() Composer 2 Thinking Enabled | 73.70 | 2026-03-19 | Unknown | Closed |
11 | ![]() DeepSeek-V4-Flash Thinking Level · Extra HighTools | 73.30 | 2026-04-24 | 284B | Free Commercial |
12 | ![]() Kimi K2.5 Thinking Enabled | 73.00 | 2026-01-27 | 1000B | Free Commercial |
13 | ![]() Claude Opus 4.6 Extended ThinkingTools | 72.00 | 2026-02-05 | Unknown | Closed |
14 | ![]() Doubao Seed 2.0 Pro Thinking EnabledTools | 71.70 | 2026-02 | Unknown | Closed |
15 | ![]() Qwen3.6-27B Thinking EnabledTools | 71.30 | 2026-04-22 | 27B | Free Commercial |
16 | ![]() DeepSeek-V4-Flash Thinking Level · HighTools | 70.20 | 2026-04-24 | 284B | Free Commercial |
17 | ![]() DeepSeek-V4-Pro Standard ModeTools | 69.80 | 2026-04-24 | 1600B | Free Commercial |
18 | ![]() DeepSeek-V4-Flash Standard ModeTools | 69.70 | 2026-04-24 | 284B | Free Commercial |
19 | ![]() Qwen3.5-397B-A17B Thinking Enabled | 69.30 | 2026-02-16 | 39.7B | Free Commercial |
20 | ![]() Qwen3.6-35B-A3B Thinking Enabled | 67.20 | 2026-04-16 | 35B | Free Commercial |
21 | ![]() Composer 1.5 Thinking Enabled | 65.90 | 2026-02-09 | Unknown | Closed |
22 | ![]() Composer 1 Thinking Enabled | 56.90 | 2025-10-29 | Unknown | Closed |