Fable-5 vs Inkling 2 Small

New Comparison

Scores are fetched live from the current database on every visit — nothing is cached.

Winner

Inkling 2 Small

Inkling 2 Small scores higher overall across category strength profiles (86 vs 64 for Fable-5); only 2 benchmarks are directly comparable.

Inkling 2 Small 86 Fable-5 63.8

Only 2 of 53 benchmarks are scored by every selected model — the verdict leans on category strength profiles and is indicative.

Benchmark Scores Comparison

Fable-5 Inkling 2 Small

Fable-5

Score: 15.0

Inkling 2 Small

594.1B params

Coverage matrix

2 models × 53 benchmarks — hover a benchmark name for its category

Full Coverage Matrix
Hides benchmarks where all selected models score identically.
Shading = score relative to the best score among the selected models: Low Mid High Missing (—) The totals row shows how many of the 53 benchmarks each model has scores for.
Benchmark Inkling 2 Small Thinking Machines Fable-5 Anthropic
GPQA Diamond 93 92.6
Automation-Bench — 29.1
SWE-bench Pro — 80
BioMysteryBench 70 —
PRBench-Finance — 55.8
Agents' Last Exam — 23.8
Toolathlon Verified — 77.9
PLawBench — 70.2
TAU3-Bench 40.4 —
ScreenSpot Pro 83.8 —
QwenReactBench — 1770
MuirBench 79.2 —
QwenSVGBench — 1690
DeepSearch QA 90.9 —
PRBench-Legal — 57.6
MMMU-Pro 81.6 —
PostTrainBench — 41.8
DSBench-FullStack — 77.2
Terminal-Bench 3.0 — 33.7
WorkSpaceBench — 68.7
Terminal-Bench Science 0.1 5.7 —
ExploitGym (6h) — 247
HLE (with tools) — 64.5
CharXiv (RQ) 85.6 —
FrontierSWE — 88.2
SkillsBench — 70.9
CoWorkBench — 75.9
SWE-bench Multilingual 81.2 —
PaperBench — 88.8
IFBench — 63.5
Humanity's Last Exam — 53.3
DSBench-Hard — 68.3
Global-MMLU-Lite 91 —
QwenSWEBench — 86.3
ExploitGym (2h) — 181
Terminal-Bench 2.1 (Terminus-2) 81.3 —
DeepSWE 72.1 70
ExploitBench — 78
HMMT Feb 26 95.8 —
Terminal Bench 2.1 — 84.6
JobBench — 57.4
$OneMillion-Bench — 55.9
QwenQoderBench — 63.1
WideSearch — 81.2
SimpleVQA 68.2 —
ProgramBench — 33
MCP-Atlas 85 —
AndroidBench — 84.5
SWE-Marathon — 33.1
Cybergym — 83.1
DeepSWE 1.1 — 70
MLS-Bench-Lite — 49.9
GDPVal-AA v2 — 1743
Covered 16/53 39/53

Scorecard

Overall 0–100 score — category strengths (50%), shared-benchmark wins (30%), coverage (20%)

Model Overall score Shared benchmark wins Coverage Strongest at Weakest at
Inkling 2 Small Winner 86 2 of 2 16 of 53 (30.0%) STEM & Reasoning Agentic Coding
Fable-5 63.8 0 of 2 39 of 53 (74.0%) STEM & Reasoning Agentic Coding

Category strengths

Where each model stands, per benchmark category — score relative to the best model, adjusted for benchmark coverage. This also ranks models when they share few benchmarks.

Category Inkling 2 Small Fable-5
Agentic Coding 21 benchmarks 100 97.1
STEM & Reasoning 5 benchmarks 100 99.2

Fully comparable benchmarks

Scored by all 2 selected models (2)

Benchmark Category Fable-5 Inkling 2 Small
GPQA Diamond stem_reasoning 92.6 93
DeepSWE coding_agent 70 72.1

Partial coverage benchmarks

Scored by at least one but not all selected models (51) — missing scores show as —

Benchmark Category Fable-5 Inkling 2 Small
Automation-Bench 1/2 models general_agent 29.1 —
SWE-bench Pro 1/2 models coding_agent 80 —
BioMysteryBench 1/2 models stem_reasoning — 70
PRBench-Finance 1/2 models general_capabilities 55.8 —
Agents' Last Exam 1/2 models general_agent 23.8 —
Toolathlon Verified 1/2 models general_agent 77.9 —
PLawBench 1/2 models general_capabilities 70.2 —
TAU3-Bench 1/2 models general_agent — 40.4
ScreenSpot Pro 1/2 models general_agent — 83.8
QwenReactBench 1/2 models coding_agent 1770 —
MuirBench 1/2 models video_understanding — 79.2
QwenSVGBench 1/2 models coding_agent 1690 —
DeepSearch QA 1/2 models general_agent — 90.9
PRBench-Legal 1/2 models general_capabilities 57.6 —
MMMU-Pro 1/2 models vision_language — 81.6
PostTrainBench 1/2 models coding_agent 41.8 —
DSBench-FullStack 1/2 models coding_agent 77.2 —
Terminal-Bench 3.0 1/2 models coding_agent 33.7 —
WorkSpaceBench 1/2 models general_agent 68.7 —
Terminal-Bench Science 0.1 1/2 models coding_agent — 5.7
ExploitGym (6h) 1/2 models cybersecurity 247 —
HLE (with tools) 1/2 models stem_reasoning 64.5 —
CharXiv (RQ) 1/2 models document_understanding — 85.6
FrontierSWE 1/2 models coding_agent 88.2 —
SkillsBench 1/2 models general_agent 70.9 —
CoWorkBench 1/2 models general_agent 75.9 —
SWE-bench Multilingual 1/2 models coding_agent — 81.2
PaperBench 1/2 models coding_agent 88.8 —
IFBench 1/2 models instruction_following 63.5 —
Humanity's Last Exam 1/2 models stem_reasoning 53.3 —
DSBench-Hard 1/2 models coding_agent 68.3 —
Global-MMLU-Lite 1/2 models multilingual — 91
QwenSWEBench 1/2 models coding_agent 86.3 —
ExploitGym (2h) 1/2 models cybersecurity 181 —
Terminal-Bench 2.1 (Terminus-2) 1/2 models coding_agent — 81.3
ExploitBench 1/2 models cybersecurity 78 —
HMMT Feb 26 1/2 models stem_reasoning — 95.8
Terminal Bench 2.1 1/2 models coding_agent 84.6 —
JobBench 1/2 models general_agent 57.4 —
$OneMillion-Bench 1/2 models general_capabilities 55.9 —
QwenQoderBench 1/2 models coding_agent 63.1 —
WideSearch 1/2 models general_agent 81.2 —
SimpleVQA 1/2 models vision_language — 68.2
ProgramBench 1/2 models coding_agent 33 —
MCP-Atlas 1/2 models general_agent — 85
AndroidBench 1/2 models coding_agent 84.5 —
SWE-Marathon 1/2 models coding_agent 33.1 —
Cybergym 1/2 models general_agent 83.1 —
DeepSWE 1.1 1/2 models coding_agent 70 —
MLS-Bench-Lite 1/2 models coding_agent 49.9 —
GDPVal-AA v2 1/2 models general_agent 1743 —

Comparison FAQ

Which model is better, Inkling 2 Small or Fable-5?

Inkling 2 Small scores higher overall across category strength profiles (86 vs 64 for Fable-5); only 2 benchmarks are directly comparable.

Which model is better at Agentic Coding, Inkling 2 Small or Fable-5?

Inkling 2 Small leads the Agentic Coding category (strength score 100/100 across 1 comparable benchmark).

How reliable is this comparison verdict?

Only 2 of 53 benchmarks are scored by every selected model — the verdict leans on category strength profiles and is indicative.

What now?

You've seen 2 models side by side — keep the loop going.

Share this comparison

The URL is the comparison — copy or share it anywhere. No tracking, no cookies.

Not the right matchup?

Swap one model out — the pre-filled form keeps the rest of the selection.

Swap Fable-5 × Swap Inkling 2 Small ×