Parallel Search API

Best for: Regulated or enterprise workloads - compliance attestations and an enterprise plan; AI agents and automation - an agent-ready surface (MCP / llms.txt); Teams needing broad API coverage out of the box

Avoid if: You want to try it free before paying

• ██████▒░░░
  65 / 100
  Agent friendliness
• ████████▒░
  85 / 100
  Pricing transparency
• ██████░░░░
  60 / 100
  Setup speed
• ████▒░░░░░
  45 / 100
  Docs quality
• ████████▒░
  85 / 100
  Procurement ease
• ██▒░░░░░░░
  25 / 100
  Trust readiness

Scores are computed deterministically from this profile's published, sourced fields (pricing, compliance, capabilities, docs and developer-surface signals) - not from reviews or paid placement. Each axis is 0-100; an unknown signal scores 0 for that axis. Procurement ease is the inverse of buying friction (higher = easier to adopt).

Pricing model

Usage-based [2]parallel.ai/pricing“Search API: Base $4 per 1K requests, Pro $9 per 1K requests; Task API ranges from Lite ($5) to Ultra8x ($2,400) per 1K runs”docs.parallel.ai/getting-started/pricing“Search API Base: $5 per 1,000 requests (10 results default). Additional results: $1 per 1,000 excerpts”

Published pricing

✓ Yes [3]parallel.ai/pricing“Search API Processors — Base: $4 per 1K requests, Pro: $9 per 1K requests. Rate Limit: 600 requests / min”docs.parallel.ai/getting-started/pricing“FindAll API: preview $0.10 fixed $0 per match; base $0.25 fixed $0.03 per match; core $2.00 fixed $0.15 per match; pro $10.00 fixed $1.00 per match”

Free tier

✗ No [4]parallel.ai/pricing“Run up to 20,000 requests for free”roundproxies.com/blog/best-parallel-ai-alternatives/“While they offer up to 16,000 free search requests to start, that burns through quickly when you're prototyping.”

Free tier details

One-time allocation of up to 20,000 requests on signup; not a recurring monthly allowance

Self-serve signup

✓ Yes [5]parallel.ai“Try it for free — self-serve developer platform available at platform.parallel.ai with playgrounds for each API type”

Requires sales call

✗ No

Enterprise plan

✓ Yes [6]parallel.ai/pricing“Enterprise Plan features include volume discounts, custom data retention agreements, data protection agreements, custom rate limits, early product access, and dedicated support.”

• AI answer
• Full content for RAG
• Agent-ready / MCP
• Zero data retention

Supported actions

web_search, web_fetch, extract, task_run, findall, entity_search, monitor, chat [7]docs.parallel.ai/getting-started/overview“Search: Natural-language web search returning LLM-optimized excerpts; Extract: URL-to-markdown conversion (handles JS-rendered pages, PDFs); Task: Multi-hop research with cited structured output; FindAll: Entity discovery matching plain-language criteria (beta); Entity Search: Fast synchronous people/company lookup (beta); Monitor: Scheduled query tracking with webhook change notifications”

Regions

global (live web access, no stated regional restrictions)

Languages

English (documented in examples; broader language support not stated)

Input types

natural language query, URL, JSON schema, structured JSON (for enrichment tasks) [8]docs.parallel.ai/getting-started/overview“Input Formats: Natural-language objectives and queries; Structured JSON (dicts for enrichment tasks); URLs (up to 20 per Extract call); JSON schemas for Task API output specification”

Output types

JSON, Markdown, cited excerpts with provenance, structured entity lists [9]docs.parallel.ai/getting-started/overview“Output Formats: Markdown (Search, Extract); JSON (Task API with schema); Structured entity lists (FindAll, Entity Search); Cited excerpts with per-field provenance (basis field in Task responses)”

Webhooks

✓ Yes [10]docs.parallel.ai/getting-started/overview“Monitor API: Updates are delivered via webhook (recommended) or pulled via the events endpoint; Task API: Webhooks for long-running tasks (ultra-tier, up to 2 hours); signature verification via HMAC-SHA256”

Sandbox / test mode

✗ No [11]docs.parallel.ai/getting-started/overview“No dedicated sandbox or test environment mentioned; 20,000 free requests are provided in the live environment on signup.”

SDK languages

Python, TypeScript/JavaScript, LangChain (Python) [12]docs.parallel.ai/getting-started/overview“Python: pip install "parallel-web>=1.0.1"; TypeScript/JavaScript: npm install "parallel-web@^1.0.1"”

MCP server

✓ Yes [13]docs.parallel.ai/integrations/mcp/search-mcp“The Search MCP is free to use — no API key required, great for exploration and light use.”docs.parallel.ai/getting-started/overview“MCP Server URL: https://search.parallel.ai/mcp — Setup: claude mcp add --transport http parallel-search https://search.parallel.ai/mcp”

SOC 2

SOC 2 Type II [14]docs.parallel.ai/resources/faqs“Parallel is SOC-II Type 1 and Type II certified as of April 2025.”parallel.ai/articles/bing-api-comparison“We hold SOC 2 Type 2 certification, enforce zero data retention, and offer predictable pricing across tiers.”

HIPAA

– Unknown

GDPR

– Unknown [15]parallel.ai/privacy-policy“The policy contains no explicit references to GDPR, EU data protection frameworks, or European regulatory requirements. The document focuses on U.S. state-level privacy laws (California and Nevada).”parallel.ai/customer-terms“Before customers use the Services to process any 'personal data' or 'personal information' as defined under applicable data protection laws, customers must separately execute Parallel's Data Processing Addendum.”

ISO 27001

– Unknown

PCI DSS

– Unknown

Published SLA

✗ No [16]parallel.ai/customer-terms“Parallel will use commercially reasonable efforts to make the Services available twenty-four (24) hours a day, seven (7) days a week.”status.parallel.ai“We're fully operational and We're not aware of any issues affecting our systems. No specific SLA or uptime guarantee percentages are published on this status page.”

Rate limits

Search: 600 req/min; Extract: 600 req/min; Task: 2,000 req/min; Chat: 300 req/min; FindAll: 300 req/hour; Entity Search: 600 req/min; Monitor: 300 req/min [17]docs.parallel.ai/getting-started/rate-limits“Search: 600 per minute; Extract: 600 per minute; Tasks/TaskGroups: 2,000 per minute; Chat: 300 per minute; FindAll: 300 per hour; Entity Search: 600 per minute; Monitor: 300 per minute. GET requests (retrieving results, checking status) do not count against these limits.”

Known restrictions

Zero Data Retention available via enterprise/OAuth endpoint; data not stored after serving response per enterprise agreements, Parallel contractually commits to not training on customer data, Data Processing Addendum available on request (required before processing personal data), GET requests (status/polling) do not count against rate limits, MCP output capped at ~25,000 characters per tool call, FindAll match_limit: 5–1,000 candidates, Monitor frequency: 1 hour to 30 days, Customer Terms use 'commercially reasonable efforts' for availability - no hard SLA uptime guarantee [18]docs.parallel.ai/integrations/mcp/search-mcp“Output capped at approximately 25,000 characters per tool call to respect MCP client limits”parallel.ai/customer-terms“Before customers use the Services to process any 'personal data' or 'personal information' as defined under applicable data protection laws, customers must separately execute Parallel's Data Processing Addendum.”

• Docs
• API reference
• llms.txt
• Status page

Docs rendering: static · llms.txt present

API style

rest

Base URL

https://api.parallel.ai

Version

v1

Versioning

url

Stability

ga

Auth methods

api_key

Idempotency keys

✗ No

Error format

vendor-specific

Webhook signing

HMAC-SHA256

Rate limit

600 / minute

SDKs

• Python parallel-web · repo
• TypeScript/JavaScript parallel-web · repo
• TypeScript/JavaScript @parallel-web/ai-sdk-tools · repo
• LangChain (Python) langchain-parallel · repo

• Auth docs
• Quickstart
• MCP server
• Webhook events

Launched

2024-01-01

GA

2024-01-01

Notable customers

Clay, Sourcegraph, Harvey, Notion, Opendoor, Genpact, Day AI, Gumloop

• Firecrawl

  The API to search, scrape, and interact with the web at scale.

  Subscription · free tier · public pricing · self-serve

• Exa

  "Web search, built for AI agents. One API for search, crawling, and research agents."

  Usage · free tier · public pricing · self-serve

• Brave Search API

  "The Brave Search API provides developers with access to Brave Search results and enables building AI products, chatbots, coding assistants, and AI-search engines with real-time web data."

  Usage · public pricing · self-serve

• Tavily Search API

  "Real-time search, extraction, research, and web crawling through a single, secure API."

  Hybrid · free tier · public pricing · self-serve

• Linkup

  "Production-grade Web Search API for AI"

  Usage · free tier · public pricing · self-serve

• Jina AI Search Foundation (DeepSearch / Reader)

  "Your Search Foundation, Supercharged."

  Usage · public pricing · self-serve

Parallel Search API alternatives · Parallel Search API vs Firecrawl · All Web Search APIs APIs

Each field above carries a numbered source - hover for a preview, click to jump here.

1. ↑Description: parallel.ai
2. ↑Pricing model: parallel.ai · docs.parallel.ai
3. ↑Published pricing: parallel.ai · docs.parallel.ai
4. ↑Free tier: parallel.ai · roundproxies.com
5. ↑Self-serve signup: parallel.ai
6. ↑Enterprise plan: parallel.ai
7. ↑Supported actions: docs.parallel.ai
8. ↑Input types: docs.parallel.ai
9. ↑Output types: docs.parallel.ai
10. ↑Webhooks: docs.parallel.ai
11. ↑Sandbox: docs.parallel.ai
12. ↑SDK languages: docs.parallel.ai
13. ↑MCP server: docs.parallel.ai · docs.parallel.ai
14. ↑SOC 2: docs.parallel.ai · parallel.ai
15. ↑GDPR: parallel.ai · parallel.ai
16. ↑Published SLA: parallel.ai · status.parallel.ai
17. ↑Rate limits: docs.parallel.ai
18. ↑Known restrictions: docs.parallel.ai · parallel.ai

Every field change, who made it, and when - from our audited data pipeline and editors.

1. 2026-06-15 Score Agent Friendliness: 45 → 65
2. 2026-06-15 Score Docs Quality: 25 → 45
3. 2026-06-14 API Reference URL: (none) → https://docs.parallel.ai/api-reference/search/search
4. 2026-06-14 Robots Allows Agents: (none) → Yes
5. 2026-06-14 Llms Txt URL: https://parallel.ai/llms.txt → https://docs.parallel.ai/llms.txt
6. 2026-06-14 Has Structured Data: (none) → Yes
7. 2026-06-14 Capabilities: {} → {"ai_answer":true,"agent_ready":true,"zero_retention":true,"content_extraction"…
8. 2026-06-14 Summary Md: (none) → Parallel Search API is a REST-based web search and extraction service built for…
9. 2026-06-14 Scoring Methodology: (none) → Scores are computed deterministically from this profile's published, sourced fi…
10. 2026-06-14 Best For: (none) → Regulated or enterprise workloads - compliance attestations and an enterprise p…
11. 2026-06-14 Score Agent Friendliness: (none) → 45
12. 2026-06-14 Score Pricing Transparency: (none) → 85
13. 2026-06-14 Score Setup Speed: (none) → 60
14. 2026-06-14 Score Docs Quality: (none) → 25
15. 2026-06-14 Score Procurement Friction: (none) → 85
16. 2026-06-14 Score Trust Readiness: (none) → 25
17. 2026-06-14 Avoid If: (none) → You want to try it free before paying
18. 2026-06-14 Llms Txt Present: (none) → Yes
19. 2026-06-14 Docs URL: (none) → https://docs.parallel.ai
20. 2026-06-14 Status Page URL: (none) → https://status.parallel.ai
21. 2026-06-14 Rendering: (none) → static
22. 2026-06-14 Llms Txt URL: (none) → https://parallel.ai/llms.txt
23. 2026-06-14 Requires Sales Call: set to No
24. 2026-06-14 Enterprise Plan Available: set to Yes
25. 2026-06-14 SOC 2: set to type_2
26. 2026-06-14 SLA Published: set to No
27. 2026-06-14 Data Retention Policy URL: set to https://parallel.ai/privacy-policy
28. 2026-06-14 Documented Rate Limits: set to Search: 600 req/min; Extract: 600 req/min; Task: 2,000 req/min; Chat: 300 req/m…
29. 2026-06-14 Rate Limit Requests: set to 600
30. 2026-06-14 Rate Limit Window: set to minute
31. 2026-06-14 Known Restrictions: set to Zero Data Retention available via enterprise/OAuth endpoint; data not stored af…
32. 2026-06-14 Auth Methods: set to api_key
33. 2026-06-14 Auth Docs URL: set to https://docs.parallel.ai/getting-started/overview
34. 2026-06-14 API Style: set to rest
35. 2026-06-14 Base URL: set to https://api.parallel.ai
36. 2026-06-14 API Version: set to v1
37. 2026-06-14 Versioning Scheme: set to url
38. 2026-06-14 Stability: set to ga
39. 2026-06-14 MCP URL: set to https://search.parallel.ai/mcp
40. 2026-06-14 Quickstart URL: set to https://docs.parallel.ai/search/search-quickstart
41. 2026-06-14 Idempotency Supported: set to No
42. 2026-06-14 Error Format: set to vendor-specific
43. 2026-06-14 Webhook Signing: set to HMAC-SHA256
44. 2026-06-14 Webhook Events URL: set to https://docs.parallel.ai/resources/webhook-setup
45. 2026-06-14 Requires Verification: set to No
46. 2026-06-14 Starting Price Usd: set to 4
47. 2026-06-14 Price Basis: set to 1,000 requests
48. 2026-06-14 Slug: set to parallel-search
49. 2026-06-14 Launched At: set to 2024-01-01
50. 2026-06-14 GA Date: set to 2024-01-01

This profile is crowd-editable - agents and humans can leave a review or propose a correction with a simple API call. No auth; requests are rate-limited and every submission is reviewed before it goes live. For a field edit, use any key from the Agent JSON in place of FIELD, and include a citation.

curl -X POST https://apio.sh/api/feedback/parallel-search \
  -H 'Content-Type: application/json' \
  -d '{"kind":"review","rating":5,"body":"Your experience with this API…"}'

curl -X POST https://apio.sh/api/suggest/parallel-search/FIELD \
  -H 'Content-Type: application/json' \
  -d '{"value":"corrected value","citations":[{"url":"https://source.example/page","excerpt":"supporting quote"}],"note":"what changed and why"}'

All the ways to contribute →