Parallel
[Menu]

Pricing

# Pay as you go

Run up to 20,000 requests for free

|                   | Task API                                                  | Search API                          | Chat API                     |
| ----------------- | --------------------------------------------------------- | ----------------------------------- | ---------------------------- |
| Inputs            | Existing structured data, Question                        | Search objective, Queries           | Question                     |
| Outputs           | Deep research, Structured enrichments                     | Ranked URLs, Excerpts               | Free text, Structured JSON   |
| Best for          | Deep research, database enrichment,or workflow automation | Web search tool calls for AI agents | Interactive chat apps and UX |
| Latency           | 5s - 30min, asynchronous                                  | <3s, synchronous                    | <5s, synchronous             |
| Basis             | Citations, reasoning, confidence, excerpts                | —                                   | Citations                    |
| Rate limits       | 2,000 requests / min                                      | 600 requests / min                  | 300 requests / min           |
| Security          | SOC2                                                      | SOC2                                | SOC2                         |
| Price per request | $0.005 - $2.4                                             | $0.004 - $0.009                     | $0.005                       |

Task API

Deep web research with structured outputs, optimized for quality & freshness

Price per request:
$0.005 - $2.4
Task playground[Task playground](https://platform.parallel.ai/play)
T
then
P
Inputs:
Existing structured dataQuestion
Outputs:
Deep researchStructured enrichments
Best for:
Deep research, database enrichment,or workflow automation
Latency:
5s - 30min, asynchronous
Basis:
Citations, reasoning, confidence, excerpts
Rate limits:
2,000 requests / min
Security:
SOC2

Search API

Ranked web URLs with long, relevant content

Price per request:
$0.004 - $0.009
Search playground[Search playground](https://platform.parallel.ai/play/search)
S
then
P
Inputs:
Search objectiveQueries
Outputs:
Ranked URLsExcerpts
Best for:
Web search tool calls for AI agents
Latency:
<3s, synchronous
Basis:
—
Rate limits:
600 requests / min
Security:
SOC2

Chat API

Fast, web-researched LLM completions

Price per request:
$0.005
Chat playground[Chat playground](https://platform.parallel.ai/play/chat)
C
then
P
Inputs:
Question
Outputs:
Free textStructured JSON
Best for:
Interactive chat apps and UX
Latency:
<5s, synchronous
Basis:
Citations
Rate limits:
300 requests / min
Security:
SOC2

## Enterprise

Built for scalable, reliable, enterprise workloads

Get a demo
D
[Get a demo](https://form.fillout.com/t/sL37Ja5wWKus)
  • - Volume discounts
  • - Custom data retention agreements
  • - Data Protection Agreement
  • - Volume discounts
  • - Custom data retention agreements
  • - Data Protection Agreement
  • - Custom rate limits
  • - Early access to new products
  • - Dedicated onboarding and technical support
  • - Custom rate limits
  • - Early access to new products
  • - Dedicated onboarding and technical support

## Allocate compute based on task complexity

Parallel lets you easily flex compute based on the complexity of your task

## Pricing breakdown

Basic information retrieval
$5.001,000 requests

### Task API

| Processor | Cost (per 1K requests) | Best for                               | Latency    | Max Fields | Basis                                      |
| --------- | ---------------------- | -------------------------------------- | ---------- | ---------- | ------------------------------------------ |
| Lite      | $5                     | Basic information retrieval            | 5s-60s     | ~2         | Citations, Reasoning                       |
| Base      | $10                    | Simple web research                    | 15s-100s   | ~5         | Citations, Reasoning                       |
| Core      | $25                    | Complex web research                   | 1min-5min  | ~10        | Citations, Reasoning, Excerpts, Confidence |
| Pro       | $100                   | Exploratory web research               | 3min-9min  | ~20        | Citations, Reasoning, Excerpts, Confidence |
| Ultra     | $300                   | Extensive deep research                | 5min-25min | ~20        | Citations, Reasoning, Excerpts, Confidence |
| Ultra2x   | $600                   | Advanced deep research with 2x compute | 5min-25min | ~25        | Citations, Reasoning, Excerpts, Confidence |
| Ultra4x   | $1,200                 | Advanced deep research with 4x compute | 8min-30min | ~25        | Citations, Reasoning, Excerpts, Confidence |
| Ultra8x   | $2,400                 | Advanced deep research with 8x compute | 8min-30min | ~25        | Citations, Reasoning, Excerpts, Confidence |

### Search API

| Processor | Cost (per 1K requests) | Best for           | Latency |
| --------- | ---------------------- | ------------------ | ------- |
| Base      | $4                     | Low latency search | 1s-3s   |
| Pro       | $9                     | Fresh search       | 45s-70s |

### Chat API

| Processor | Cost (per 1K requests) | Best for                                | Latency | Basis     |
| --------- | ---------------------- | --------------------------------------- | ------- | --------- |
| Speed     | $5                     | Low-latency, web-based chat completions | <5s     | Citations |
Processor

Cost (per 1K requests)

Best for

Latency

Max Fields

Basis

Lite

$5

Basic information retrieval

5s-60s

~2

Citations, Reasoning

Base

$10

Simple web research

15s-100s

~5

Citations, Reasoning

Core

$25

Complex web research

1min-5min

~10

Citations, Reasoning, Excerpts, Confidence

Pro

$100

Exploratory web research

3min-9min

~20

Citations, Reasoning, Excerpts, Confidence

Ultra

$300

Extensive deep research

5min-25min

~20

Citations, Reasoning, Excerpts, Confidence

Ultra2x

$600

Advanced deep research with 2x compute

5min-25min

~25

Citations, Reasoning, Excerpts, Confidence

Ultra4x

$1,200

Advanced deep research with 4x compute

8min-30min

~25

Citations, Reasoning, Excerpts, Confidence

Ultra8x

$2,400

Advanced deep research with 8x compute

8min-30min

~25

Citations, Reasoning, Excerpts, Confidence

Start building

## Scale with predictable cost

Parallel APIs are priced per request, not per token. You always know the exact cost of a query before you run it.

Developer platform
P
[Developer platform](https://platform.parallel.ai)Get a demo
D
[Get a demo](https://form.fillout.com/t/sL37Ja5wWKus)

[Get up to $5K in free credits]

Qualified startups can receive up to $5K in free credits from Parallel

Apply now[Apply now](https://form.fillout.com/t/cNxJPKmh7eus)

[Use pre-committed spend on Parallel]

Use pre-committed AWS spend on Parallel via the AWS marketplace

Get in touch[Get in touch](https://form.fillout.com/t/x5mAateBUnus)

Highest quality at every price point

State of the art across several benchmarks

BrowseCompDeepResearch BenchWISER-SearchWISER-AtomicSimpleQA
Performance comparison proving Parallel delivers the best enterprise deep research API for ChatGPT and AI agents with 48% accuracy vs competitors' 14% max across Model and Accuracy (%). Multi-hop research benchmark shows Parallel's structured AI agent deep research outperforms GPT-4, Claude, Exa, and Perplexity. Enterprise-ready structured deep research API with MCP server integration.

### About the benchmark

This benchmark[benchmark]($https://openai.com/index/browsecomp/), created by OpenAI, contains 1,266 questions requiring multi-hop reasoning, creative search formulation, and synthesis of contextual clues across time periods. Results are reported on a random sample of 100 questions from this benchmark. Learn more in our blog[blog]($https://parallel.ai/blog/introducing-parallel).

### Methodology

  • - Dates: All measurements were made between 08/08/2025 and 08/13/2025.
  • - Configurations: For all competitors, we report the highest numbers we were able to achieve across multiple configurations of their APIs. The exact configurations are below.
    • - GPT-5: high reasoning, high search context, default verbosity
    • - Exa: Exa Research Pro
    • - Anthropic: Claude Opus 4.1
    • - Perplexity: Sonar Deep Research reasoning effort high

## Highest quality at every price point

State of the art across several benchmarks

### Browsecomp (LP)

| Category   | Accuracy (%) |
| ---------- | ------------ |
| Ultra8x    | 58           |
| Ultra      | 45           |
| Pro        | 34           |
| GPT-5      | 41           |
| Exa        | 14           |
| Anthropic  | 7            |
| Perplexity | 6            |

### About the benchmark

This benchmark[benchmark]($https://openai.com/index/browsecomp/), created by OpenAI, contains 1,266 questions requiring multi-hop reasoning, creative search formulation, and synthesis of contextual clues across time periods. Results are reported on a random sample of 100 questions from this benchmark. Learn more in our blog[blog]($https://parallel.ai/blog/introducing-parallel).

### Methodology

  • - Dates: All measurements were made between 08/08/2025 and 08/13/2025.
  • - Configurations: For all competitors, we report the highest numbers we were able to achieve across multiple configurations of their APIs. The exact configurations are below.
    • - GPT-5: high reasoning, high search context, default verbosity
    • - Exa: Exa Research Pro
    • - Anthropic: Claude Opus 4.1
    • - Perplexity: Sonar Deep Research reasoning effort high

### DeepResearch Bench (LP)

| Category | Win Rate (%) |
| -------- | ------------ |
| Ultra8x  | 82           |
| Ultra    | 74           |
| GPT-5    | 66           |

### About the benchmark

This benchmark[benchmark]($https://github.com/Ayanami0730/deep_research_bench) contains 100 expert-level research tasks designed by domain specialists across 22 fields, primarily Science & Technology, Business & Finance, and Software Development. It evaluates AI systems' ability to produce rigorous, long-form research reports on complex topics requiring cross-disciplinary synthesis. Results are reported from the subset of 50 English-language tasks in the benchmark. Learn more in our blog[blog]($https://parallel.ai/blog/introducing-parallel).

### Methodology

  • - Dates: All measurements were made between 08/08/2025 and 08/13/2025.
  • - Win Rate: Calculated by comparing RACE[RACE]($https://github.com/Ayanami0730/deep_research_bench) scores in direct head-to-head evaluations.
  • - Configurations: For all competitors, we report results for the highest numbers we were able to achieve across multiple configurations of their APIs. The exact GPT-5 configuration is high reasoning, high search context, and high verbosity.
  • - Excluded API Results: Exa Research Pro (0% win rate), Claude Opus 4.1 (0% win rate), and Perplexity Sonar Deep Research (6% win rate).

### Search MCP Benchmark (LP) 

| Series    | Model                        | Cost (CPM) | Accuracy (%) |
| --------- | ---------------------------- | ---------- | ------------ |
| Parallel  | GPT 4.1 w/ Prll Search MCP   | 21         | 74.9         |
| Parallel  | o4 mini w / Prll Search MCP  | 90         | 82.14        |
| Parallel  | o3 w / Prll Search MCP       | 192        | 80.61        |
| Parallel  | sonnet 4 w / Prll Search MCP | 92         | 78.57        |
| Native    | GPT 4.1 w / Native Search    | 27         | 70           |
| Native    | o4 mini w / Native Search    | 190        | 77           |
| Native    | o3 w / Native Search         | 351        | 79.08        |
| Native    | sonnet 4 w / Native Search   | 122        | 68.83        |
| Exa       | GPT 4.1 w/ Exa Search MCP    | 40         | 58.67        |
| Exa       | o4 mini w/ Exa Search MCP    | 199        | 61.73        |
| Exa       | o3 w/ Exa Search MCP         | 342        | 56.12        |
| Exa       | sonnet 4 w/ Exa Search MCP   | 140        | 67.13        |

CPM: USD per 1000 requests. Cost is shown on a Linear scale.

### About the benchmark

This benchmark, created by Parallel, blends WISER-Fresh and WISER-Atomic. WISER-Fresh is a set of 76 queries requiring the freshest data from the web, generated by Parallel with o3 pro. WISER-Atomic is a set of 120 hard real-world business queries, based on use cases from Parallel customers. Read our blog here[here]($https://parallel.ai/blog/search-api-benchmark).

### Distribution

40% WISER-Fresh
60% WISER-Atomic

### Evaluation

The Parallel Search API was evaluated by comparing three different web search solutions (Parallel MCP server, Exa MCP server/tool calling, LLM native web search) across four different LLMs (GPT 4.1, o4-mini, o3, Claude Sonnet 4).

### WISER-Atomic

| Series   | Model          | Cost (CPM) | Accuracy (%) |
| -------- | -------------- | ---------- | ------------ |
| Parallel | Core           | 25         | 77           |
| Parallel | Base           | 10         | 75           |
| Parallel | Lite           | 5          | 64           |
| Others   | o3             | 45         | 69           |
| Others   | 4.1 mini low   | 25         | 63           |
| Others   | gemini 2.5 pro | 36         | 56           |
| Others   | sonar pro high | 16         | 64           |
| Others   | sonar low      | 5          | 48           |

CPM: USD per 1000 requests. Cost is shown on a Log scale.

### About the benchmark

This benchmark, created by Parallel, contains 121 questions intended to reflect real-world web research queries across a variety of domains. Read our blog here[here]($https://parallel.ai/blog/parallel-task-api).

### Steps of reasoning

50% Multi-Hop questions
50% Single-Hop questions

### Distribution

40% Financial Research
20% Sales Research
20% Recruitment
20% Miscellaneous

### SimpleQA

| Series   | Model            | Cost (CPM) | Accuracy (%) |
| -------- | ---------------- | ---------- | ------------ |
| Parallel | Lite             | 5          | 92           |
| Parallel | Base             | 10         | 94           |
| Parallel | Core             | 25         | 94           |
| Others   | o3 high          | 56         | 92           |
| Others   | 4.1 mini high    | 30         | 88           |
| Others   | gemini 2.5 flash | 35         | 91           |
| Others   | sonar            | 8          | 81           |
| Others   | sonar pro        | 13         | 84           |

CPM: USD per 1000 requests. Cost is shown on a Log scale.

### About the benchmark

This benchmark contains 4,326 questions focused on short, fact-seeking queries across a variety of domains. Read our blog here[here]($https://parallel.ai/blog/parallel-task-api).

### Steps of reasoning

100% Single-Hop questions

### Distribution

36% Culture
20% Science and Technology
16% Politics
28% Miscellaneous