How many tracked prints does the SWE-bench Verified family have?

SWE-bench Verified leaderboard — the family board.

9 TRACKED VARIANTS · 4 FIRST-CLASS BOARDS

One family, 9 tracked prints: SWE-bench Pro, swe_bench_bash and SWE-bench Multimodal are cited standalone and keep live boards of their own. Scoring configs and splits of the base board render as sections of this page, each under its own honest label.

● LIVE DATA147 MODELS ON THIS BOARDLEADER CLAUDE OPUS 5 AT 96

First-class boards in this familyCITED STANDALONE · EACH KEEPS ITS OWN LIVE BOARD

#BoardModels scored

01SWE-bench VerifiedTHIS PAGE147

02SWE-bench Pro90

03swe_bench_bash40

04SWE-bench Multimodal15

The board — SWE-bench VerifiedTOP 100 OF 147 MEASURED · BEST SCORE PER MODEL, DEFAULT VARIANT

#ModelVendorScoreVLast seen

01Claude Opus 5Anthropic962026-07-24

02Claude Mythos 5Anthropic95.52026-06-10

03Claude Fable 5Anthropic952026-08-23

04Claude Mythos PreviewAnthropic93.92026-08-23

05GPT-5.6 LunaOpenAI932026-07-30

06Ember-1fireworks92.22026-10-08

07Claude Opus 4.8Anthropic88.62026-08-23

08Claude Opus 4.7Anthropic87.62026-08-23

09Claude Sonnet 5Anthropic85.22026-08-23

10Claude Opus 4.5Anthropic81.52026-09-25

11Claude Opus 4.6Anthropic80.82026-08-23

12Gemini 3.1 ProGoogle80.62026-08-24

13DeepSeek-V4-ProDeepSeek80.62026-08-23

14MiniMax M3MiniMax80.52026-08-23

15Qwen3.7 MaxAlibaba80.42026-08-23

16Inkling-SmallThinking Machines80.22026-08-23

17Kimi K2.6Moonshot80.22026-08-23

18North-Mini-Code-1.0Cohere80.22026-08-23

19MiniMax M2.5MiniMax80.22026-09-11

20GPT-5.2OpenAI802026-09-25

21MiniMax M2.7MiniMax79.92026-07-30

22Claude Sonnet 4.6Anthropic79.62026-08-23

23DeepSeek-V4-FlashDeepSeek792026-07-30

24MiMo V2.5 ProXiaomi78.92026-08-23

25Qwen3.6 PlusAlibaba78.82026-08-23

26MiMo V2 ProXiaomi782026-08-23

27Hy3-previewTencent782026-08-23

28Gemini 3 ProGoogle782026-09-25

29Gemini 3 Flash PreviewGoogle782026-09-25

30GPT-5.1 Codex MaxOpenAI77.92026-06-15

31GLM-5Z.ai77.82026-09-25

32Qwen3.7 Plus PreviewAlibaba77.72026-08-23

33InklingThinking Machines77.62026-09-09

34Mistral Medium 3.5 128BMistral77.62026-08-23

35Muse SparkMeta77.42026-08-23

36MiMo V2.5 Pro FP4 DFlashXiaomi77.42026-08-29

37Qwen3.6 27BAlibaba77.22026-08-23

38Claude Sonnet 4.5Anthropic77.22026-09-25

39Kimi K2.5Moonshot76.82026-09-25

40Seed 2.0ByteDance76.52026-09-04

41Qwen3.5 397B A17BAlibaba76.42026-08-23

42GPT-5.1 InstantOpenAI76.32026-08-23

43GPT-5.1OpenAI76.32026-09-25

44GLM-5.1Z.ai76.22026-06-06

45Muse GlimmerMeta762026-08-23

46Qwen3 Max (Reasoning)Alibaba75.32026-08-23

47Gemini 3.5 Flash LiteGoogle752026-07-30

48Qwen3.5 27BAlibaba752026-08-23

49GPT-5OpenAI74.92026-09-25

50MiMo V2 OmniXiaomi74.82026-08-23

51GPT-5 CodexOpenAI74.52026-08-23

52Claude Opus 4.1Anthropic74.52026-08-23

53Step 3.5 FlashStepFun74.42026-08-23

54MiniMax M2.1MiniMax742026-08-23

55GLM-4.7Z.ai73.82026-08-23

56GPT-5.1 CodexOpenAI73.72026-09-25

57MAI-Thinking-1Microsoft73.52026-08-23

58Qwen3.6 35B A3BAlibaba73.42026-08-23

59MiMo V2 FlashXiaomi73.42026-08-23

60Claude Haiku 4.5Anthropic73.32026-08-23

61DeepSeek-V3.2DeepSeek73.12026-09-25

62DeepSeek-V3.2-SpecialeDeepSeek73.12026-08-23

63Seed 1.8ByteDance72.92026-09-09

64GPT-5.2 CodexOpenAI72.82026-09-01

65Claude Sonnet 4Anthropic72.72026-09-09

66Claude Opus 4Anthropic72.52026-08-23

67Devstral 2Mistral72.22026-09-08

68Qwen3.5 122B A10BAlibaba722026-08-23

69MAI-Code-1-FlashMicrosoft71.62026-08-23

70Kimi K2 ThinkingMoonshot71.32026-09-01

71MiMo V2.5Xiaomi712026-07-30

72Laguna-XS-2.1poolside70.92026-08-23

73Grok Code Fast 1SpaceXAI70.82026-08-23

74Nemotron 3 Ultra 550B A55BNVIDIA70.72026-09-01

75Solar Pro4Upstage70.62026-08-23

76Claude 3.7 SonnetAnthropic70.32026-09-01

77Qwen3.5 35B A3BAlibaba702026-08-23

78Qwen3 Coder 480B A35B InstructAlibaba69.62026-08-23

79Qwen3 Coder PlusAlibaba69.62026-06-15

80MiniMax M2MiniMax69.42026-09-11

81Kimi K2 InstructMoonshot69.22026-08-23

82o3OpenAI69.12026-09-01

83o4-miniOpenAI68.12026-09-01

84Devstral Small 2Mistral682026-06-15

85GLM-4.6Z.ai682026-08-23

86DeepSeek-V3.2-ExpDeepSeek67.82026-08-23

87Gemini 2.5 ProGoogle67.22026-09-03

88DeepSeek-V3.1DeepSeek662026-08-23

89GLM-4.5Z.ai64.22026-08-23

90Nemotron 3 SuperNVIDIA63.12026-08-11

91Devstral MediumMistral61.62026-08-23

92Gemini 2.5 FlashGoogle60.42026-09-01

93GPT-5 miniOpenAI59.82026-09-25

94GLM-4.7-FlashZ.ai59.22026-08-23

95Nemotron CORTEXANVIDIA58.22026-05-01

96GLM-4.5-AirZ.ai57.62026-08-23

97Gemma 4 26B A4BGoogle57.42026-08-11

98Granite 4.2 30BIBM572026-08-31

99MiniMax M1 80KMiniMax562026-08-23

100MiniMax M1 40KMiniMax55.62026-08-23

◆ IV — INDEPENDENTLY VERIFIED · ◆ AA — AGGREGATOR · ◆ VA — VENDOR · ◇ SA — SOURCE-ATTRIBUTED

Variants of this boardSCORING CONFIGS & SPLITS — NOT CITED STANDALONE, SO THEY LIVE HERE

SWE-bench Multilingual

VARIANT77 models · best score per model

#ModelVendorScoreVLast seen

01Claude Opus 5.5Anthropic93.92026-09-28

02Claude Mythos 5Anthropic92.22026-09-01

03Sonnet 5.5Anthropic90.32026-10-07

04Claude Opus 5Anthropic89.52026-09-22

05Claude Mythos 5.1Anthropic89.12026-09-01

06Claude Fable 5.1Anthropic89.12026-09-22

07Claude Mythos PreviewAnthropic87.32026-08-23

08Claude Fable 5Anthropic86.62026-09-01

09Claude Opus 4.8Anthropic84.42026-08-23

10Haiku 5.5Anthropic83.72026-10-08

11Hy4Tencent82.92026-08-31

12Qwen3.8 Flash NextAlibaba812026-08-27

13Claude Opus 4.7Anthropic80.52026-06-21

14Laguna S 2.1poolside78.52026-08-23

15Qwen3.7 MaxAlibaba78.32026-08-23

Other tracked prints

THIN BOARDSunder 5 published models each — listed, never hidden

PrintLeaderScoreModels

swe_bench_liteClaude Sonnet 458.33

SWE-bench Verified (hard subset)Claude Sonnet 4.545.33

SWE-Bench (OpenHands)Nemotron 3 Super 120B A12B59.52

swe_bench_testClaude 3 Opus3.81

THE FAMILY BOARD · EVERY PRINT LABELED · EVERY SCORE SOURCE-ATTRIBUTED ON ITS OWN LIVE BOARDAll leaderboards The frontier board Models by vendor