Fable 5.1 vs GPT-6.1 Sol

Fable 5.1 leads, winning 3 of 5 directly-comparable benchmarks against GPT-6.1 Sol.

Head-to-head record: Fable 5.1 3 · 2 GPT-6.1 Sol

Fable 5.1vsGPT-6.1 Sol

Fable 5.1

$10/$50 per Mtok

Fable 5

$10/$50 per Mtok

Opus 5.5

$4/$20 per Mtok

Opus 5

$5/$25 per Mtok

Opus 4.8

$5/$25 per Mtok

Sonnet 5.5

$2/$10 per Mtok

Sonnet 5

$3/$15 per Mtok

GPT-6 Astra

$10/$50 per Mtok

GPT-6.1 Sol

$2/$10 per Mtok

GPT-5.6 Sol

$5/$30 per Mtok

Grok 4.7

$2/$6 per Mtok

GPT-5.5

$5/$30 per Mtok

Composer 2.5

$0.5/$2.5 per Mtok

Opus 4.7

$5/$25 per Mtok

Gemini 3.1 Pro

$2/$12 per Mtok

Mythos Preview

DeepSWE

—

69.7%

74.2%

68.8%

59.0%

71.0%

54.0%

74.1%

75.2%

72.7%

71.0%

67.0%

—

—

11.8%

—

SWE-bench Pro

81.2%

80.3%

89.9%

79.2%

69.2%

81.3%

63.2%

—

—

64.6%

—

59.4%

—

64.3%

54.2%

77.8%

SWE-bench Verified

—

95.0%

—

97.0%

88.6%

—

79.6%

—

—

96.2%

—

82.6%

79.6%

82.0%

78.8%

—

SWE-bench ML

89.1%

—

93.9%

89.5%

—

90.3%

—

—

—

—

—

77.8%

79.8%

80.5%

—

—

FrontierCode 1.1

50.3%

—

54.4%

48.0%

—

46.2%

42.4%

53.3%

—

47.5%

—

—

—

—

—

—

T-Bench 4.0

57.9%

44.5%

64.8%

53.9%

23.6%

61.8%

12.4%

58.2%

58.2%

37.3%

37.6%

—

—

—

—

—

HLE (tools)

65.6%

63.8%

67.7%

63.6%

—

64.5%

54.9%

57.2%

—

—

—

—

—

—

—

—

BrowseComp

—

—

—

90.8%

84.3%

—

84.7%

—

—

90.4%

default

92.2%

ultra

—

84.4%

—

79.8%

85.9%

86.9%

MCP-Atlas

—

—

—

85.8%

82.2%

—

—

—

—

—

—

75.3%

—

79.1%

78.2%

—

AutomationBench

—

—

42.5%

—

—

44.7%

10.7%

41.4%

36.1%

—

—

—

—

—

—

—

OSWorld 2.1

80.7%

partial

42.8%

strict

—

81.8%

partial

48.7%

strict

74.0%

partial

37.2%

strict

—

80.1%

partial

43.5%

strict

57.0%

partial

25.6%

strict

—

—

—

—

—

—

—

—

—

AA-LCR v1.1

85.3%

82.3%

84.7%

79.3%

77.7%

82.7%

82.0%

80.7%

83.0%

84.0%

76.7%

84.3%

—

78.7%

82.0%

—

Blueprint-Bench 2

—

38.6%

—

—

14.5%

—

—

—

—

—

—

36.2%

—

—

26.5%

—

Finance Agent v2

58.9%

56.3%

58.6%

58.6%

53.9%

58.1%

53.9%

53.5%

52.0%

53.8%

52.3%

51.8%

—

51.5%

43.0%

—

GDPval-AA v2.1

1756

1613

1866

1724

1456

1839

1465

1542

1575

1611

1712

1353

—

1356

794

—

GDPpdf

—

29.8%

—

—

22.5%

—

—

—

—

30.7%

—

26.0%

—

—

16.7%

—

Legal Agent

—

13.3%

8.3%

11.7%

10.4%

10.0%

5.8%

—

—

—

—

2.1%

—

—

0.0%

—

CyberGym

—

83.8%

—

—

78.8%

—

—

—

—

84.5%

—

81.8%

—

73.1%

—

83.1%

ExploitBench

—

78.0%

91.0%

70.0%

40.0%

80.0%

—

—

—

73.5%

—

47.9%

—

—

—

69.0%

BioMysteryBench

—

—

89.3%

human solvable

50.0%

human difficult

91.4%

human solvable

51.8%

human difficult

—

—

84.9%

human solvable

39.4%

human difficult

—

—

—

—

—

—

—

—

90.3%

human solvable

44.1%

human difficult

HealthBench Prof.

62.1%

66.0%

65.6%

59.8%

56.9%

69.2%

57.8%

64.7%

64.2%

60.5%

56.7%

49.5%

—

—

—

64.7%

GPQA Diamond

—

94.1%

—

93.7%

93.6%

—

—

96.0%

—

94.6%

—

93.6%

—

94.2%

94.3%

94.6%

Chartography

44.8%

no tools

88.4%

with tools

—

64.4%

no tools

89.0%

with tools

29.8%

no tools

83.4%

with tools

—

61.6%

no tools

90.2%

with tools

—

—

—

—

—

—

—

—

—

—

MMMLU

—

—

—

—

—

—

—

—

—

—

—

83.2%

—

91.5%

92.6%

—