AWS StrandsPerplexityDecision matchup

Strands Decider 2Bvs.pplx-decider-v1.1-27b

Which model makes better decisions?

pplx-decider-v1.1-27b does better in independent tests, while Strands Decider 2B is much smaller.

Independent tests

Strands Decider 2B0 of 1

pplx-decider-v1.1-27bahead1 of 1

Jev Decision Index

Popularity

Strands Decider 2B1

pplx-decider-v1.1-27bahead42

out of 100

Weights and access

Strands Decider 2Bopen, local

pplx-decider-v1.1-27bopen, local

Size

Strands Decider 2Bsmaller2 B

pplx-decider-v1.1-27b26.09 B

Parameters

Pick Strands Decider 2B for

smaller

Pick pplx-decider-v1.1-27b for

Jev Decision Indexmore popular

Shared tests

Independent evaluators and leaderboards only, the newest shared snapshot per evaluator. With 95 % ranges a lead counts when they do not overlap; for hit rates, when it holds in a 95 % test. Without any uncertainty figure at most “slightly ahead”.

Jev Decision Index

pplx-decider-v1.1-27b ahead

pplx-decider-v1.1-27b is clearly ahead in 27 of 51 subtests and slightly in 19, Strands Decider 2B in 0 clearly and slightly in 2. 3 are level.

Home appliances · Decision Index 0.3

Strands Decider 2B0.0

pplx-decider-v1.1-27b80.7

88 tasks, as of October 7, 2026

POP909 · Decision Index 0.3

Strands Decider 2B6.9

pplx-decider-v1.1-27b57.1

2,000 tasks, as of October 7, 2026

GSM8K · Decision Index 0.3

Strands Decider 2B24.7

pplx-decider-v1.1-27b76.6

2,638 tasks, as of October 7, 2026

Show all 51 subtests

Subtest	Strands Decider 2B	pplx-decider-v1.1-27b	Verdict
Home appliances · Decision Index 0.3	0.0	80.7	pplx-decider-v1.1-27b ahead
POP909 · Decision Index 0.3	6.9	57.1	pplx-decider-v1.1-27b ahead
GSM8K · Decision Index 0.3	24.7	76.6	pplx-decider-v1.1-27b ahead
SATA-Bench · Decision Index 0.3	9.8	28.2	pplx-decider-v1.1-27b ahead
CRUXEval · Decision Index 0.3	33.3	85.6	pplx-decider-v1.1-27b ahead
MMLU-Pro · Decision Index 0.3	31.1	71.0	pplx-decider-v1.1-27b ahead
API-Bank · Decision Index 0.3	39.6	85.8	pplx-decider-v1.1-27b ahead
ChessBench · Decision Index 0.3	9.2	17.7	pplx-decider-v1.1-27b ahead
BBH · Decision Index 0.3	46.5	84.0	pplx-decider-v1.1-27b ahead
MMLU · Decision Index 0.3	51.0	87.7	pplx-decider-v1.1-27b ahead
BPoMP · Decision Index 0.3	56.0	95.2	pplx-decider-v1.1-27b ahead
GPQA Diamond · Decision Index 0.3	33.2	56.1	pplx-decider-v1.1-27b ahead
New Yorker · Decision Index 0.3	45.5	75.6	pplx-decider-v1.1-27b ahead
HellaSwag · Decision Index 0.3	57.4	93.2	pplx-decider-v1.1-27b ahead
WinoGrande · Decision Index 0.3	57.4	92.4	pplx-decider-v1.1-27b ahead
CLadder · Decision Index 0.3	54.4	79.3	pplx-decider-v1.1-27b ahead
PhishNChips · Decision Index 0.3	50.8	73.2	pplx-decider-v1.1-27b ahead
Decision Index all-suite calibration accuracy	54.1	76.1	pplx-decider-v1.1-27b ahead
MuSR · Decision Index 0.3	51.1	70.7	pplx-decider-v1.1-27b ahead
HoVer · Decision Index 0.3	54.1	74.1	pplx-decider-v1.1-27b ahead
ARC-Challenge · Decision Index 0.3	71.1	97.0	pplx-decider-v1.1-27b ahead
cfcolor · Decision Index 0.3	50.7	64.5	pplx-decider-v1.1-27b ahead
When2Call · Decision Index 0.3	59.6	73.9	pplx-decider-v1.1-27b ahead
Humicroedit · Decision Index 0.3	49.7	61.2	pplx-decider-v1.1-27b ahead
BFCL · Decision Index 0.3	81.4	97.2	pplx-decider-v1.1-27b ahead
Habermas · Decision Index 0.3	38.0	42.0	pplx-decider-v1.1-27b ahead
ARC-Easy · Decision Index 0.3	90.5	99.3	pplx-decider-v1.1-27b ahead
Decision Index confidence-above-95% share (lower is better)	0.0	2.0	Strands Decider 2B slightly ahead
ACOS · Decision Index 0.3	6.0	28.2	pplx-decider-v1.1-27b slightly ahead
Decision Index 0.3 · Full Decision Index score	22	63	pplx-decider-v1.1-27b slightly ahead
Decision Index 0.3 · Private new-domains component	20	56	pplx-decider-v1.1-27b slightly ahead
Decision Index 0.3 · Public breadth skill score	23	62	pplx-decider-v1.1-27b slightly ahead
Decision Index 0.3 · Private same-skills component	23	61	pplx-decider-v1.1-27b slightly ahead
Decision Index 0.3 · Public index skill score	25	62	pplx-decider-v1.1-27b slightly ahead
iSarcasmEval · Decision Index 0.3	29.6	71.0	pplx-decider-v1.1-27b slightly ahead
RAGTruth · Decision Index 0.3	35.7	83.1	pplx-decider-v1.1-27b slightly ahead
Amazon ESCI · Decision Index 0.3	26.0	60.4	pplx-decider-v1.1-27b slightly ahead
VAST · Decision Index 0.3	40.4	76.7	pplx-decider-v1.1-27b slightly ahead
ANLI · Decision Index 0.3	39.9	74.3	pplx-decider-v1.1-27b slightly ahead
Decision Index all-suite Brier score (lower is better)	55.2	33.5	pplx-decider-v1.1-27b slightly ahead
Decision Index all-suite expected calibration error (lower is better)	4.4	7.1	Strands Decider 2B slightly ahead
Decision Index 0.3 · Public index raw score	44	72	pplx-decider-v1.1-27b slightly ahead
CLINC150 · Decision Index 0.3	64.7	91.5	pplx-decider-v1.1-27b slightly ahead
BRIGHT · Decision Index 0.3	36.8	50.8	pplx-decider-v1.1-27b slightly ahead
FinEntity · Decision Index 0.3	72.7	95.4	pplx-decider-v1.1-27b slightly ahead
NLI4CT · Decision Index 0.3	65.8	84.9	pplx-decider-v1.1-27b slightly ahead
ToolRet · Decision Index 0.3	60.5	67.8	pplx-decider-v1.1-27b slightly ahead
ContractNLI · Decision Index 0.3	73.0	78.4	pplx-decider-v1.1-27b slightly ahead
HLE · Decision Index 0.3	14.0	16.2	level
RouterBench · Decision Index 0.3	78.9	79.8	level
BANKING77 · Decision Index 0.3	80.4	80.6	level

Source · latest October 7, 2026

All data side by side

	Strands Decider 2B	pplx-decider-v1.1-27b
Kind	Adapter	Standalone
Developer	AWS Strands	Perplexity
Status	Open weights	Open weights
Released	October 5, 2026	October 6, 2026
Weights	open	open
Runs locally	yes	yes
Parameters	2 B	26.09 B
License	Apache-2.0	Apache-2.0; the public Hugging Face repository is not access-gated
Price	n/a	Perplexity states $0.02 per 1M input tokens.
Popularity (out of 100)	1	42

More comparisons with Strands Decider 2B and pplx-decider-v1.1-27b

All decision models compared