Find the right AI for you

Search by name or tell us what you want to do
API servicesChoose how to access modelsAI modelsCapabilities, specifications, and pricingAI toolsFind useful tools by taskMCP directoryConnect tools, data, and workflowsAgent skillsFind reusable skills for AI agentsBenchmarksCompare model performanceOnline toolsConvert accounts, format JSON, estimate costs and test APIsModel timelineExplore model releasesUse casesStart with a practical task
EnterSearch · ↑↓Choose a resultEscClose
Submit中文
MODEL DEVELOPER

Perplexity

Explore official APIs and base prices for web-connected answers, text embeddings and multimodal decision models.

Official documentation
On this page

Explore this developer’s model families

Listed families
3
Versions
13
Field review
0

13 entries use official model documentation and 0 use model-card metadata. Explore specifications, uses, prices and release records by version.

Filter these versions in the directory →View coverage →

About the organization

A company profile, legal entity, ownership and finances require separate review. An API brand entry is not a complete company profile.

No organization profile yet. See the official information and specifications for the models below.

Model families

PPLX Decider →4 versions

Models for judgments, classification and probability scores on text and images.

PPLX Embed →4 versions

Embeddings for semantic retrieval over independent text and shared document context.

Sonar →5 versions

Web-connected answer models in the Agent API.

Billing & payment details

Perplexity API Platform

Listed currency
USD
Billing model
Prepaid / account credit / Usage-based

API usage draws from prepaid credits without a subscription. Credit cards are added through Stripe; adding a payment method alone does not charge it. Optional auto-reload uses a balance threshold. API keys stop when credits run out and resume after top-up. APIs bill separately by tokens, search requests, tool calls or sandbox sessions. Consumer subscriptions are separate from API credits. As of 2026-10-29.

Official sources ↗Official sources ↗

Billing, currencies and benefits apply to the products and dates shown.API usage and subscription benefits are separate. Listed and settlement currencies may differ; actual payment depends on your account conditions.

Developer model prices

Prices are per million text tokens, matched by currency, processing mode and input tier. Caching, tools, media and taxes may cost extra. These are not per-request or subscription totals. A listed price does not guarantee that a model is still available.

Prices observed 2026-10-31 · Matching input tier 7 versions /1 developers.Other matching versions: 6 versions have no price under these conditions. Quoted currency does not imply payment support. Check each model’s own input limit before choosing a long-input tier.

Chart shows the first 7 records; matching values and sources are listed below. Input and output prices are separate. No currency conversion is applied.

USD/ million text tokens, Standard
Model / developerInputOutputCache readTier / source

sonar →Perplexity · Listed by provider

USD 0.25USD 2.5USD 0.0625

No extra input tier · All day

No end date provided

Official prices ↗

Last observed2026-10-29

pplx-decider-flash-v1.2 →Perplexity · Listed by provider

USD 0.01USD 0Not provided

No extra input tier · All day

No end date provided

Official prices ↗

Last observed2026-10-29

pplx-decider-v1.2 →Perplexity · Listed by provider

USD 0.02USD 0Not provided

No extra input tier · All day

No end date provided

Official prices ↗

Last observed2026-10-29

pplx-embed-context-v1-0.6b →Perplexity · Listed by provider

USD 0.008Not providedNot provided

No extra input tier · All day

No end date provided

Official prices ↗

Last observed2026-10-29

pplx-embed-context-v1-4b →Perplexity · Listed by provider

USD 0.05Not providedNot provided

No extra input tier · All day

No end date provided

Official prices ↗

Last observed2026-10-29

pplx-embed-v1-0.6b →Perplexity · Listed by provider

USD 0.004Not providedNot provided

No extra input tier · All day

No end date provided

Official prices ↗

Last observed2026-10-29

pplx-embed-v1-4b →Perplexity · Listed by provider

USD 0.03Not providedNot provided

No extra input tier · All day

No end date provided

Official prices ↗

Last observed2026-10-29

All quotes, future prices and conditions (7 records)
  • sonar · Standard · No extra input tier · All day

    Input USD 0.25 / output USD 2.5 / cache read USD 0.0625, per million text tokens.

    EffectiveSource not listed· EndedSource not listed· Superseded by a newer observationNot yet · First observed2026-10-21/ latest2026-10-29

    Perplexity’s own Sonar models in the Agent API, with standard processing. Search, sandbox and other tools are separate; these prices are not extended to flex or priority.

    Check sources ↗
  • pplx-decider-flash-v1.2 · Standard · No extra input tier · All day

    Input USD 0.01 / output USD 0, per million text tokens.

    EffectiveSource not listed· EndedSource not listed· Superseded by a newer observationNot yet · First observed2026-10-21/ latest2026-10-29

    Only input tokens are charged; output is explicitly free. Standard models share state across a question group, while Flash counts state again for each question.

    Check sources ↗
  • pplx-decider-v1.2 · Standard · No extra input tier · All day

    Input USD 0.02 / output USD 0, per million text tokens.

    EffectiveSource not listed· EndedSource not listed· Superseded by a newer observationNot yet · First observed2026-10-21/ latest2026-10-29

    Only input tokens are charged; output is explicitly free. Standard models share state across a question group, while Flash counts state again for each question.

    Check sources ↗
  • pplx-embed-context-v1-0.6b · Standard · No extra input tier · All day

    Input USD 0.008, per million Text tokens.

    EffectiveSource not listed· EndedSource not listed· Superseded by a newer observationNot yet · First observed2026-10-21/ latest2026-10-29

    Charged per million input tokens. Embedding dimensions are not output tokens. Shared-document batch terms follow the API documentation.

    Check sources ↗
  • pplx-embed-context-v1-4b · Standard · No extra input tier · All day

    Input USD 0.05, per million Text tokens.

    EffectiveSource not listed· EndedSource not listed· Superseded by a newer observationNot yet · First observed2026-10-21/ latest2026-10-29

    Charged per million input tokens. Embedding dimensions are not output tokens. Shared-document batch terms follow the API documentation.

    Check sources ↗
  • pplx-embed-v1-0.6b · Standard · No extra input tier · All day

    Input USD 0.004, per million Text tokens.

    EffectiveSource not listed· EndedSource not listed· Superseded by a newer observationNot yet · First observed2026-10-21/ latest2026-10-29

    Charged per million input tokens. Embedding dimensions are not output tokens. Shared-document batch terms follow the API documentation.

    Check sources ↗
  • pplx-embed-v1-4b · Standard · No extra input tier · All day

    Input USD 0.03, per million Text tokens.

    EffectiveSource not listed· EndedSource not listed· Superseded by a newer observationNot yet · First observed2026-10-21/ latest2026-10-29

    Charged per million input tokens. Embedding dimensions are not output tokens. Shared-document batch terms follow the API documentation.

    Check sources ↗

Versions at a glance

  • sonar →

    perplexity/sonar

    Uses / inputs & outputs
    Chat

    Text → Text

    Access and licensing
    Official API

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Context window: No single value specified

    Listed by provider

  • pplx-decider-flash-v1.2 →

    pplx-decider-flash-v1.2

    Uses / inputs & outputs
    Specialist models · Image understanding

    Text / Image → Scores

    Access and licensing
    Official API

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Max input: 256K Token

    Listed by provider

  • pplx-decider-v1.2 →

    pplx-decider-v1.2

    Uses / inputs & outputs
    Specialist models · Image understanding

    Text / Image → Scores

    Access and licensing
    Official API

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Max input: 256K Token

    Listed by provider

  • pplx-embed-context-v1-0.6b →

    pplx-embed-context-v1-0.6b

    Uses / inputs & outputs
    Embeddings

    Text → Vectors

    Access and licensing
    Official API

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Max input: 32K Token

    Listed by provider

  • pplx-embed-context-v1-4b →

    pplx-embed-context-v1-4b

    Uses / inputs & outputs
    Embeddings

    Text → Vectors

    Access and licensing
    Official API

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Max input: 32K Token

    Listed by provider

  • pplx-embed-v1-0.6b →

    pplx-embed-v1-0.6b

    Uses / inputs & outputs
    Embeddings

    Text → Vectors

    Access and licensing
    Official API

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Max input: 32K Token

    Listed by provider

  • pplx-embed-v1-4b →

    pplx-embed-v1-4b

    Uses / inputs & outputs
    Embeddings

    Text → Vectors

    Access and licensing
    Official API

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Max input: 32K Token

    Listed by provider

  • llama-3.1-sonar-huge-128k-online →

    llama-3.1-sonar-huge-128k-online

    Uses / inputs & outputs
    Chat

    Text → Text

    Access and licensing
    Access not specified

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Context window: No single value specified

    Current status unknown

  • llama-3.1-sonar-large-128k-online →

    llama-3.1-sonar-large-128k-online

    Uses / inputs & outputs
    Chat

    Text → Text

    Access and licensing
    Access not specified

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Context window: No single value specified

    Current status unknown

  • llama-3.1-sonar-small-128k-online →

    llama-3.1-sonar-small-128k-online

    Uses / inputs & outputs
    Chat

    Text → Text

    Access and licensing
    Access not specified

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Context window: No single value specified

    Current status unknown

  • pplx-decider-v1.1-27b →

    pplx-decider-v1.1-27b

    Uses / inputs & outputs
    Specialist models · Image understanding

    Text / Image → Scores

    Access and licensing
    Access not specified

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Context window: No single value specified

    Current status unknown

  • pplx-decider-v1-27b →

    pplx-decider-v1-27b

    Uses / inputs & outputs
    Specialist models · Image understanding

    Text / Image → Scores

    Access and licensing
    Access not specified

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Context window: No single value specified

    Current status unknown

  • sonar-reasoning →

    sonar-reasoning

    Uses / inputs & outputs
    Chat

    Text → Text

    Access and licensing
    Access not specified

    Openness not specified · Weight license needs review · Commercial weight use unknown

    Limits & data status
    Context window: No single value specified

    Current status unknown

Release & lifecycle history

Release & lifecycle events

8 sourced events by month, including releases, deprecations and retirements. One event can cover several versions. These are release events in the current catalog, rather than newly added listings.

Monthly event counts
  • 2025-02 · 1records
  • 2025-12 · 1records
  • 2026-09 · 1records
  • 2026-10 · 5records

8 events; showing the latest 8 below. The chart includes all documented events.Open the full timeline →

  1. API launchDecisions API: pplx-decider-v1.1-27b replaces v1 →

    New Decisions API models: pplx-decider-v1.1-27b.

  2. API launchDecisions API: pplx-decider-v1.2 and pplx-decider-flash-v1.2 →

    New Decisions API models: pplx-decider-v1.2, pplx-decider-flash-v1.2.

  3. API launchNew: Decisions API →

    New Decisions API models: pplx-decider-v1-27b.

  4. Retirementpplx-decider-v1.1-27b replaced →

    The earlier decision model has been superseded. Its API ID routes to v1.2; the exact transition date is unknown.

  5. Retirementpplx-decider-v1-27b replaced →

    The earlier decision model has been superseded. Its API ID routes to v1.2; the exact transition date is unknown.

  6. Deprecation noticeLegacy Sonar Chat Completions support ended →

    The legacy product API is no longer supported. Synchronous and streaming calls are progressively rewritten as Agent requests; asynchronous calls stop. This does not retire the current Agent Sonar models.

  7. Retirementsonar-reasoning removed →

    This API model was removed. The official recommendation is to migrate to Sonar Reasoning Pro.

  8. RetirementLegacy Sonar online models removed →

    Three legacy Sonar online API models were retired.

References & scope

Each version also lists its own specification and license references. A US developer brand. Hosted models’ original developer countries, deployment locations and supported regions are recorded separately. Catalog as of 2026-10-31.

Find APIs, tools & benchmarks

Find APIs by model family, then check each service for the exact model, region and supported interfaces. Tools and reports link to specific versions.