Fable-5
Anthropic
Parameters
—
Architecture
—
Released
—
License
—
About
Fable-5 (w/ fallback) - proprietary model by Anthropic, used as comparison model in DeepSeek-V4-Pro-0813 benchmark table.
Benchmark Scores
| Benchmark | Score | Date |
|---|---|---|
|
Terminal Bench 2.1
coding_agent
|
93.35%
|
18.08.2026 |
|
SWE-bench Pro
coding_agent
|
100.00%
|
18.08.2026 |
|
DeepSWE 1.1
coding_agent
|
93.66%
|
18.08.2026 |
|
MLS-Bench-Lite
coding_agent
|
100.00%
|
18.08.2026 |
|
PaperBench
coding_agent
|
93.27%
|
18.08.2026 |
|
AndroidBench
coding_agent
|
100.00%
|
18.08.2026 |
|
QwenSWEBench
coding_agent
|
100.00%
|
18.08.2026 |
|
QwenQoderBench
coding_agent
|
100.00%
|
18.08.2026 |
|
QwenReactBench
coding_agent
|
100.00%
|
18.08.2026 |
|
QwenSVGBench
coding_agent
|
73.75%
|
18.08.2026 |
|
CoWorkBench
general_agent
|
100.00%
|
18.08.2026 |
|
JobBench
general_agent
|
77.06%
|
18.08.2026 |
|
SkillsBench
general_agent
|
93.67%
|
18.08.2026 |
|
Automation-Bench
general_agent
|
41.59%
|
18.08.2026 |
|
Toolathlon Verified
general_agent
|
100.00%
|
18.08.2026 |
|
WideSearch
general_agent
|
90.02%
|
18.08.2026 |
|
HLE (with tools)
stem_reasoning
|
100.00%
|
18.08.2026 |
|
GPQA Diamond
stem_reasoning
|
96.79%
|
18.08.2026 |
|
Humanity's Last Exam
stem_reasoning
|
96.97%
|
18.08.2026 |
|
IFBench
instruction_following
|
67.89%
|
18.08.2026 |
|
$OneMillion-Bench
general_capabilities
|
53.61%
|
18.08.2026 |
|
PLawBench
general_capabilities
|
79.02%
|
18.08.2026 |
|
PRBench-Legal
general_capabilities
|
100.00%
|
18.08.2026 |
|
PRBench-Finance
general_capabilities
|
78.26%
|
18.08.2026 |
|
Cybergym
general_agent
|
89.88%
|
13.08.2026 |
|
DeepSWE
coding_agent
|
96.29%
|
13.08.2026 |
|
DSBench-FullStack
coding_agent
|
100.00%
|
13.08.2026 |
|
DSBench-Hard
coding_agent
|
92.59%
|
13.08.2026 |
|
Terminal-Bench 3.0
coding_agent
|
97.00%
|
28.08.2026 |
|
ProgramBench
coding_agent
|
43.54%
|
28.08.2026 |
|
FrontierSWE
coding_agent
|
100.00%
|
28.08.2026 |
|
SWE-Marathon
coding_agent
|
65.51%
|
28.08.2026 |
|
PostTrainBench
coding_agent
|
97.95%
|
28.08.2026 |
|
ExploitGym (2h)
cybersecurity
|
82.67%
|
28.08.2026 |
|
ExploitGym (6h)
cybersecurity
|
82.77%
|
28.08.2026 |
|
ExploitBench
cybersecurity
|
100.00%
|
28.08.2026 |
|
Agents' Last Exam
general_agent
|
62.26%
|
28.08.2026 |
|
GDPVal-AA v2
general_agent
|
95.19%
|
28.08.2026 |
|
WorkSpaceBench
general_agent
|
62.50%
|
18.08.2026 |