Skip to main content
BenchLM
Data

Data as of November 13, 2026 · How the score is built

Perplexity Decider v1 27B

CurrentOpen WeightNon-Reasoning

Decision readingPerplexity Decider v1 27B is tracked, but not publicly ranked yet. The profile exposes 12 sourced benchmark rows and leaves unsupported fields blank until a published record exists.

Email me when its price, rank or availability changes.

Capability

Unranked

field median 49.4Not eligible for a public rank

Price

$0.020input / $0 output

input median $0.60blended $0.010

Speed

Not measured

field median 92 tok/sTime to first token not measured

Context

262144 tokenstokens

field median 256,000Maximum output length is tracked separately

Decision snapshot

How much published evidence this model has, where that evidence is strongest, and what to validate before choosing it. It has no public rank, so no field position is shown.

Strongest published evidence

Published rows are visible, but no category has enough eligible evidence for a comparative rank.

Validate before choosing

12 published rows leave some tracked benchmark slots empty. Independent runtime speed has not been measured.

Source-linked · 12 displayable benchmark rows

Benchmark ledger

Every result says who ran the test. For this model, Perplexity ran all 12. External signals opens by default. The tick marks the best source-verified result in the catalog; provisional leaders do not set the reference.

Overall rank
Not ranked yet
No benchmark-run or independent results yet
Reported by Perplexity
12
Shown with that label
Run by a benchmark
0
None yet
Independent runs
0
None yet
External signals12 rows
External signals benchmark results, best verified comparison, who ran each test, and whether it counts toward a rank
WinoGrande (Perplexity panel)83.30%

Best verified: Jev 1.13.0 · 90.70%

Gap7.4 behindReported by PerplexityShown only
FinancialPhraseBank (Perplexity panel)84.18%

Best verified: Perplexity Decider v1 27B · 84.18%

GapHighest listedReported by PerplexityShown only
RAGTruth (Perplexity panel)88.80%

Best verified: Perplexity Decider v1 27B · 88.80%

GapHighest listedReported by PerplexityShown only
JudgeBench (Perplexity panel)78.29%

Best verified: Jev 1.13.0 · 78.57%

Gap0.3 behindReported by PerplexityShown only
BBH (Perplexity panel)82.80%

Best verified: Jev 1.13.0 · 94.27%

Gap11.5 behindReported by PerplexityShown only
JevBench public hard (Perplexity panel)70.30%

Best verified: Jev 1.13.0 · 73.27%

Gap3 behindReported by PerplexityShown only
TabFact (Perplexity panel)90.60%

Best verified: Perplexity Decider v1 27B · 90.60%

GapHighest listedReported by PerplexityShown only
ContractNLI (Perplexity panel)80.78%

Best verified: Qwen3.8-27B · 80.78%

GapHighest listedReported by PerplexityShown only
Circa (Perplexity panel)89.20%

Best verified: Perplexity Decider v1 27B · 89.20%

GapHighest listedReported by PerplexityShown only
Belebele (Perplexity panel)94.00%

Best verified: Jev 1.13.0 · 95.00%

Gap1 behindReported by PerplexityShown only
TruthfulQA binary (Perplexity panel)85.40%

Best verified: Jev 1.13.0 · 92.00%

Gap6.6 behindReported by PerplexityShown only
Perplexity Decision Panel85.71%

Best verified: Perplexity Decider v1 27B · 85.71%

GapHighest listedReported by PerplexityShown only

Outlined bars are lab-reported. Filled bars are benchmark-run or independent. Bars run 0–100; the dark tick marks the best source-verified value.

Lineage

The sequence follows explicit supersedes links. Each score is estimated for that model; a relative can inform a sparse estimate but never sets a floor, so a newer release can score below an earlier one. Scores and prices remain blank when the corresponding public row or first-party rate is unavailable.

  1. Oct 1, 2026 · you are here

    Perplexity Decider v1 27B

    Not publicly ranked · $0.02 / $0

Decision-system

Spec sheet

Each documented value carries its source. Missing fields stay visible as not sourced or not published, rather than disappearing from the page.

API model ID
pplx-decider-v1-27bPerplexity Decisions API documentation
Maximum output
Not sourced yet
Knowledge cutoff
Not sourced yet
Input modalities
text, imagePerplexity Decisions API documentation
Output modalities
typed JSONPerplexity Decisions API documentation
Parameters
Not sourced yet
Availability
Perplexity Decisions API · Apache 2.0 open weightsPerplexity Decider released model card
Cloud regions
Not tracked yet
Lifecycle
Current
API capabilities
Typed yes/no probabilities, choices, and rubric scores; up to 128 questions per request, 255 choices, or 10 score levels.Perplexity Decisions API documentation
Prompt caching
Not documented in the pricing recordPerplexity Decisions API pricing
Self-host
Open weights available; hardware estimate not sourced
Rate limits
Not tracked yet

How to read this profile

The visual layer above carries the decisions. These notes preserve the model, ranking, coverage, and family context behind the numbers.

We track Perplexity Decider v1 27B, but the public leaderboard excludes this profile until enough non-generated benchmark coverage is available. Only published rows appear above.

Perplexity Decider v1 27B is a open weight model with a 262144 tokens context window. No explicit reasoning mode is documented in this profile.

Perplexity released this Qwen3.8-27B fine-tune under Apache 2.0 on October 1, 2026. The published 11-task decision panel reports 85.71% accuracy over 7,210 rows, measured through the Perplexity API. Those provider-reported results have separate display-only keys and do not establish a general model rank. The released config specifies 262,144 positions; API requests must stay below 262,144 input tokens across state, images, and questions. The model card describes approximately 49 GiB of weights plus working memory for local CUDA inference. Since Perplexity released pplx-decider-v1.2 on October 16, 2026, API requests that name this model ID run pplx-decider-v1.2 and bill at its $0.02 rate; the open weights are unchanged.

12 of 919 tracked benchmark slots currently have displayable evidence. Missing categories stay blank.

Last updated November 13, 2026. Runtime fields remain blank until a sourced snapshot exists.

Deployment options

Self-host and provider-specific paths stay separate from benchmark evidence so operating constraints are visible before a score becomes the whole decision.

Published weights are available, but BenchLM does not yet have a sourced parameter and VRAM profile for this exact model. Hardware cost estimates stay unavailable until that sizing record is complete.

Estimate VRAM from known parameters

Questions

How does Perplexity Decider v1 27B perform overall in AI benchmarks?

Perplexity Decider v1 27B has 12 source-displayable benchmark rows, but it does not qualify for a public overall rank. The available rows remain visible by category without being converted into a site-wide score. Missing evidence stays blank instead of being estimated from an earlier model.

Is Perplexity Decider v1 27B open source?

Perplexity Decider v1 27B is an open-weight model from Perplexity. Its weights can be downloaded for local or hosted deployment, subject to the published license. Open weight does not automatically mean open source: training data and training code may remain private, and commercial restrictions can still apply.

Does Perplexity Decider v1 27B have full benchmark coverage on BenchLM?

No. Perplexity Decider v1 27B currently has 12 source-displayable rows. The profile exposes published, non-generated evidence and leaves missing categories blank until an exact evaluation is available. Coverage describes how much was measured; it is not a penalty added to an individual benchmark result.

What is the context window size of Perplexity Decider v1 27B?

Perplexity Decider v1 27B has a documented context window of 262144 tokens. That figure is the maximum combined prompt and retained-conversation space reported for this exact model; it is not the maximum output length. The profile keeps output limits separate because providers often publish those limits independently.

Compare Perplexity Decider v1 27B with every tracked model1032 comparisons