Perplexity: Sonar Deep Research

Public pricingIntelligence 79/100Medium memoryDeep thinking

Perplexity: Sonar Deep Research is a text model for reasoning and problem solving. It combines deep reasoning and planning with a 128K tokens context window and a balanced-cost profile. Use it for reasoning, analysis, and hard problem solving when quality, speed, and cost matters. It is a practical choice for teams that need reliable output, flexible deployment, and room to scale.

Input

$2.00/1M

Output

$8.00/1M

Cached

$0.50/1M

Batch

$1.00/1M

Calculate your Sonar Deep Research bill.

Set your workload — see cost at your exact volume.

What would Sonar Deep Research cost you?

Adjust the workload to see your monthly bill.

1,00010,00050,000250,0001M10M

Technical specifications

Sonar Deep Research at a glance.

Memory

128,000

tokens

Max reply

8,192

tokens

Memory tier

Medium

a long report or a codebase file

Tokenizer

—

Released

Mar 2025

Training cutoff

Oct 2024

Availability

Public pricing

Status

active

What it can do

Capabilities & limits.

  • Understands images
  • Deep step-by-step thinking
  • Uses tools / calls functions
  • Strict JSON output
  • Streams replies
  • Fine-tunable on your data

When to pick Sonar Deep Research

  • Multi-step reasoning, research agents, or hard math.

When to look elsewhere

  • Your workload involves images — pick a vision-capable model instead.
  • You need tool-use / function calling for agent workflows.

Alternatives

Similar models to compare.

FAQ

Sonar Deep Research — the questions we see most.

Pricing, capabilities, alternatives — generated from the same data that powers the calculator above.

Get instant answers from our AI agent

At a typical workload of 50,000 conversations a month with 1,500-token prompts and 800-token replies, Sonar Deep Research costs roughly $470 per month. Input is $2.00 /1M tokens and output is $8.00 /1M tokens.
Sonar Deep Research has a 128,000-token context window (medium memory — a long report or a codebase file). That means you can fit about 24,000 words of input and history in a single call.
Beyond text generation, Sonar Deep Research supports deep step-by-step reasoning, strict JSON output. It streams replies by default.
Sonar Deep Research was released in March 2025, with training data cut off around October 2024.
Models in a similar class include Sonar Reasoning Pro, Gemini 3.1 Pro Preview, Gemini 3.1 Pro Preview Custom Tools. The "Similar models" section below this FAQ links into each.

Still unsure?

Compare Sonar Deep Research against 100+ other models.

Open the full wizard — pick a use case, set your usage, and see side-by-side monthly costs in under a minute.

Open the comparison wizard