Model comparison

Claude Fable 5 vs DeepSeek-V3

Claude Fable 5 is the stronger model overall, scoring 66.3 to 39.2 on the Noometry Index. DeepSeek-V3 costs 49× less per token, which makes it the better buy when Claude Fable 5's lead doesn't matter for your workload.

Last verified October 10, 2026. 28 shared benchmarks.

Claude Fable 5 Anthropic

66.3

Rank #5 Confirmed

DeepSeek-V3

DeepSeek

39.2

Rank #157 Confirmed

Summary

They share 28 benchmarks with published results for both. Claude Fable 5 scores higher in 8 categories and DeepSeek-V3 in 0 categories; 8 gaps are clear of the uncertainty.
The widest gap is in math, where Claude Fable 5 leads 88.4 to 30.4.
The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 100% for Claude Fable 5 and 37.8% for DeepSeek-V3.
DeepSeek-V3 is cheaper at $0.24 / $0.90 per million input/output tokens, against $10 / $50 for Claude Fable 5.
Claude Fable 5 accepts more context: 1M tokens versus 164K.
DeepSeek-V3 has downloadable open weights; the other is API-only.

Side by side

Claude Fable 5 and DeepSeek-V3 specifications
	Claude Fable 5	DeepSeek-V3
Provider	Anthropic	DeepSeek
Noometry Index	66.3	39.2
Released	2026-06-07	2024-12-26
Weights	Proprietary	Open
Context window	1M	164K
Max output	128K	164K
Input $ / M tokens	$10	$0.24
Output $ / M tokens	$50	$0.90
Results tracked	61	60

Sponsored placements are available on pages like this one. Advertise on Noometry

Category by category

Coding Claude Fable 5 leads

Claude Fable 5: 67.0 (#4), DeepSeek-V3: 41.9 (#99)

Coding benchmarks
Benchmark	Claude Fable 5	DeepSeek-V3
SciCode	61%	35.8%
WeirdML	91.9%	36.1%
LMArena Coding	1519	1368
DeepSWE	69.9%	—
FrontierCode	53.5%	—
Aider Polyglot	—	55.1%
LMArena WebDev	1626	—
FrontierSWE	47%	—
BigCodeBench Instruct	—	50%
LiveBench Coding	—	70.9%
MirrorCode	63.9%	—
BigCodeBench Complete	—	62.2%
ALE-Bench	2,041	—
HumanEval+	—	86.6%
MBPP+	—	73%

Agentic & Tool Use Not comparable

Claude Fable 5: 54.1 (#2), DeepSeek-V3: —

Agentic & Tool Use benchmarks
Benchmark	Claude Fable 5	DeepSeek-V3
APEX-Agents	63.6%	—
Remote Labor Index	16.1%	—
τ²-bench Banking	39.7%	—
PostTrainBench	41.8%	—
GBAEval	74.5%	—
GDP.pdf	30%	—
LMArena Search	1230	—
METR Time Horizons	—	49.6%
Vending-Bench 2	5,680	—

Reasoning Claude Fable 5 leads

Claude Fable 5: 76.8 (#4), DeepSeek-V3: 20.0 (#226)

Reasoning benchmarks
Benchmark	Claude Fable 5	DeepSeek-V3
SimpleBench	81.9%	27.2%
Kagi LLM Benchmark	91.4%	52.3%
CritPt	28.6%	0%
LMArena Hard Prompts	1508	1365
DTBench	98.4%	64.8%
LMCA	61.1%	15.5%
Epoch Capabilities Index	163.6	135.95
ARC-AGI-2	89.2%	—
NYT Connections (extended)	92.7%	—
ARC-AGI-1	98.5%	—
Chess Puzzles	41%	—
EnigmaEval	39.3%	—
EBR-Bench	39.5%	—
LiveBench Reasoning	—	65.8%
Mystery Game Puzzles	52%	—
LiveBench Data Analysis	—	60.9%
Surface Evolver Bench	95%	—
Bench to the Future 3	0.13	—
BIG-Bench Hard	—	87.5%
ForecastBench	—	59.1
HellaSwag	—	88.9%
LiveBench	—	66.9%
PIQA	—	84.7%
WinoGrande	—	85.2%

Math Claude Fable 5 leads

Claude Fable 5: 88.4 (#5), DeepSeek-V3: 30.4 (#208)

Math benchmarks
Benchmark	Claude Fable 5	DeepSeek-V3
OTIS Mock AIME 2024-2025	100%	37.8%
LMArena Math	1519	1373
FrontierMath (Tiers 1-3)	87%	—
FrontierMath Tier 4	90.2%	—
ProofBench	95%	—
Omni-MATH	—	40.3%
LiveBench Math	—	73.5%
MATH Level 5	—	75.5%
FrontierMath (Feb 2025 set)	—	1.7%
FrontierMath Erdős	0%	—

Knowledge Claude Fable 5 leads

Claude Fable 5: 62.3 (#22), DeepSeek-V3: 37.5 (#147)

Knowledge benchmarks
Benchmark	Claude Fable 5	DeepSeek-V3
GPQA Diamond	85.9%	67.6%
LMArena Expert	1534	1352
SimpleQA Verified	70.7%	—
MMLU-Pro	—	72.3%
Confabulations	—	26.1%
Vectara Hallucination Rate	—	6.1%
GPQA (HELM)	—	53.8%
ARC (AI2) Challenge	—	95.3%
MMLU	—	87.2%
TriviaQA	—	82.9%

Multimodal Not comparable

Claude Fable 5: 45.6 (#14), DeepSeek-V3: —

Multimodal benchmarks
Benchmark	Claude Fable 5	DeepSeek-V3
LMArena Vision	1326	—
Blueprint-Bench 2	38.6%	—
Furniture Assembly	35.8%	—
LMArena Document	1496	—

Multilingual Claude Fable 5 leads

Claude Fable 5: 57.6 (#8), DeepSeek-V3: 48.8 (#138)

Multilingual benchmarks
Benchmark	Claude Fable 5	DeepSeek-V3
LMArena Non-English	1483	1358
LMArena Chinese	1546	1390
LMArena French	1509	1386
LMArena German	1486	1374
LMArena Japanese	1507	1333
LMArena Korean	1493	1321
LMArena Russian	1507	1372
LMArena Spanish	1503	1358

Instruction Following Claude Fable 5 leads

Claude Fable 5: 78.8 (#7), DeepSeek-V3: 73.0 (#125)

Instruction Following benchmarks
Benchmark	Claude Fable 5	DeepSeek-V3
LMArena Instruction Following	1503	1345
LiveBench Instruction Following	—	81.5%
IFEval	—	83.2%

Long Context Claude Fable 5 leads

Claude Fable 5: 46.3 (#22), DeepSeek-V3: 34.0 (#248)

Long Context benchmarks
Benchmark	Claude Fable 5	DeepSeek-V3
LMArena Longer Query	1510	1353
Fiction.LiveBench	—	50%

Writing & Preference Claude Fable 5 leads

Claude Fable 5: 76.1 (#5), DeepSeek-V3: 57.6 (#124)

Writing & Preference benchmarks
Benchmark	Claude Fable 5	DeepSeek-V3
LMArena Text	1491	1375
LMArena Creative Writing	1494	1364
EQ-Bench Creative Writing	1943	1472
LMArena Multi-Turn	1505	1389
Short-Story Creative Writing	—	77%
WildBench	—	83%
EQ-Bench 4	1340	—
LiveBench Language	—	49.1%

Frequently asked questions

Is Claude Fable 5 better than DeepSeek-V3?

Which is cheaper, Claude Fable 5 or DeepSeek-V3?

DeepSeek-V3 is cheaper. It lists at $0.24 per million input tokens and $0.90 per million output tokens; Claude Fable 5 lists at $10 and $50.

Is Claude Fable 5 or DeepSeek-V3 better for coding?

Claude Fable 5 scores higher on coding benchmarks: 67.0 versus 41.9 in the Noometry coding category.

Which has the bigger context window?

Claude Fable 5 does, with 1M tokens against 164K.

How many benchmarks do Claude Fable 5 and DeepSeek-V3 share?

28 benchmarks have published results for both models. Claude Fable 5 has 61 scored results on Noometry and DeepSeek-V3 has 60.