Fastest in the ≥95% shortlist
Exa instant is fastest among configurations scoring at least 95% accuracy: 398ms average search time, 97.7% accuracy and $7.00 per 1,000 queries.
independent · open source code · public dataset sample · 2026
Most Accurate Web Search API (2026): Comparison & Benchmark
Most accurate web search API in this benchmark: Exa fast, 99.3% accuracy, 652ms average search time and $7.00 per 1,000 queries.
Compare the best web search APIs from 11 providers, including Exa, Tavily, Brave, Parallel and Firecrawl, for AI agents, LLM apps and RAG. The same 300 questions are used for every API. Compare accuracy, speed and cost per 1,000 queries with official API docs below.
Benchmark last measured: .
[01] primary comparison
Best Web Search APIs Ranked by Accuracy
19 configurations from 11 providers, ordered by measured answer accuracy. Speed and query price describe the same configuration shown in each row.
| Rank | Search API | Accuracy | Average search time | Cost per 1,000 queries | Official docs |
|---|---|---|---|---|---|
| 1 | Exa fastPOST /search type=fast | 99.3% | 652ms | $7.00Up to 10 results; extra results and summaries add cost.Pricing ↗ | Official docs ↗ |
| 2 | Exa instantPOST /search type=instant | 97.7% | 398ms | $7.00Up to 10 results; extra results and summaries add cost.Pricing ↗ | Official docs ↗ |
| 3 | Perplexity SearchPOST /search search_context_size=low | 97.3% | 1.38s | $5.00Search API · POST /searchPricing ↗ | Official docs ↗ |
| 4 | Linkup fastPOST /v1/search depth=fast · outputType=searchResults | 96.7% | 1.57s | $5.00depth=fast · searchResultsPricing ↗ | Official docs ↗ |
| 5 | SERP (RapidAPI)GET google-search74.p.rapidapi.com limit=10 | 96.0% | 751ms | $3.00Pro overage $0.003/request. Recorded 1 September 2026; confirm current pricing.Pricing ↗ | Official docs ↗ |
| 6 | FirecrawlPOST /v2/search | 95.3% | 510ms | $5.00Standard-plan top-up rate; subscription required. 2 credits for 10 results.Pricing ↗ | Official docs ↗ |
| 7 | Brave Search llm-contextPOST /res/v1/llm/context count=10 | 94.0% | 601ms | $5.00Search plan · LLM ContextPricing ↗ | Official docs ↗ |
| 8 | Parallel basicPOST /v1/search mode=basic | 93.3% | 1.68s | $5.00Up to 10 results; extra results/excerpts add cost.Pricing ↗ | Official docs ↗ |
| 9 | Brave Search webGET /res/v1/web/search count=10 · result_filter=web | 93.3% | 630ms | $5.00See linked pricing for this configuration.Pricing ↗ | Official docs ↗ |
| 10 | Nimble standardPOST /v2/search search_depth=standard · full_content=false · focus=general | 93.0% | 861ms | $5.00search_depth=standard · full_content=false. Recorded 1 September 2026; confirm current pricing.Pricing ↗ | Official docs ↗ |
| 11 | Tavily advancedPOST /search search_depth=advanced | 93.0% | 4.29s | $16.002 credits · $0.008 PAYGPricing ↗ | Official docs ↗ |
| 12 | TinyFishGET api.search.tinyfish.ai free · 30 req/min cap | 92.0% | 2.62s | $0.00Free search: 30 requests/minute, 500/hour.Pricing ↗ | Official docs ↗ |
| 13 | Linkup standardPOST /v1/search depth=standard · outputType=searchResults | 92.0% | 2.55s | $5.00depth=standard · searchResultsPricing ↗ | Official docs ↗ |
| 14 | You highlights corePOST /v1/search extraction_mode=highlights · knowledge=core | 92.0% | 889ms | $5.00highlights · knowledge=core · same Web Search price assumptionPricing ↗ | Official docs ↗ |
| 15 | You highlightsPOST /v1/search extraction_mode=highlights | 90.7% | 628ms | $5.00highlights included in Web SearchPricing ↗ | Official docs ↗ |
| 16 | Tavily basicPOST /search search_depth=basic | 87.7% | 1.88s | $8.001 credit · $0.008 PAYGPricing ↗ | Official docs ↗ |
| 17 | Parallel fastPOST /v1/search mode=fast | 86.0% | 942ms | $1.00Up to 10 results; extra results/excerpts add cost.Pricing ↗ | Official docs ↗ |
| 18 | Nimble litePOST /v2/search search_depth=lite · full_content=false · focus=general | 75.3% | 3.15s | $1.10search_depth=lite · full_content=falsePricing ↗ | Official docs ↗ |
| 19 | Parallel turboPOST /v1/search mode=turbo | 71.3% | 348ms | $1.00Up to 10 results; extra results/excerpts add cost.Pricing ↗ | Official docs ↗ |
Prices are published rates for the displayed configuration, before free credits and volume discounts. Search API charges cover requests; model tokens and additional scraping add to application cost. SERP (RapidAPI) and Nimble standard use recorded pricing from 1 September 2026; confirm current rates with the linked provider. Full pricing and free-tier comparison →
Fastest and Cheapest Search APIs with at Least 95% Accuracy
6 tested configurations meet this threshold. We use 95% as a practical shortlist filter; every result remains visible in the full ranking above.
Cheapest in the ≥95% shortlist
SERP (RapidAPI) has the lowest listed query price among configurations scoring at least 95% accuracy: $3.00 per 1,000 queries, 96.0% accuracy and 751ms average search time.
Compare the listed plan and overage basis before estimating a monthly bill.
Choose the accuracy target your application needs, then compare speed and price within that group. The threshold is a selection aid; small score differences should be checked on your own queries.
How We Measure Search Accuracy
Each API receives the same 300 company-news questions and returns up to 10 results. A fixed extraction model answers from the returned titles and snippets; a separate judge checks the answer against human-labelled ground truth. Accuracy is the percentage of questions answered correctly. The code and public dataset sample are linked below, with a held-out scoring set.
The primary ranking uses answer accuracy, then AR@5 to order equal accuracy scores. Average search time measures the API call. Benchmark measurement dates stay separate from this page’s editorial update and published pricing.
Answer quality and retrieved evidence
Accuracy scores the answer produced by the extraction model. AR@1 checks whether the first returned snippet contains the ground-truth answer; AR@5 checks whether any of the first five snippets contains it. Recall measures the evidence retrieved, while accuracy measures the answer built from that evidence.
See answer-bearing recall (AR@1 and AR@5) for every API
| Search API | Accuracy | AR@1 | AR@5 |
|---|---|---|---|
| Exa fast | 99.3% | 95.0% | 99.3% |
| Exa instant | 97.7% | 80.0% | 97.3% |
| Perplexity Search | 97.3% | 91.7% | 98.3% |
| Linkup fast | 96.7% | 79.0% | 94.7% |
| SERP (RapidAPI) | 96.0% | 78.0% | 95.0% |
| Firecrawl | 95.3% | 77.7% | 96.7% |
| Brave Search llm-context | 94.0% | 81.0% | 94.7% |
| Parallel basic | 93.3% | 55.0% | 92.0% |
| Brave Search web | 93.3% | 79.3% | 91.7% |
| Nimble standard | 93.0% | 86.3% | 94.3% |
| Tavily advanced | 93.0% | 65.0% | 92.7% |
| TinyFish | 92.0% | 74.3% | 90.7% |
| Linkup standard | 92.0% | 67.3% | 90.3% |
| You highlights core | 92.0% | 67.3% | 89.7% |
| You highlights | 90.7% | 72.0% | 89.7% |
| Tavily basic | 87.7% | 82.0% | 87.7% |
| Parallel fast | 86.0% | 44.3% | 79.0% |
| Nimble lite | 75.3% | 64.0% | 75.3% |
| Parallel turbo | 71.3% | 45.3% | 66.0% |
Open source code ↗ · Public dataset sample ↗ · Full search benchmark →
Supporting Developer Search Accuracy
100 developer questions, ranked by task completion. These scores measure a developer agent completing documentation tasks.
Developer search accuracy results
| Rank | Vendor | Endpoint & configuration | Official docs | Task completion |
|---|---|---|---|---|
| 1 | Perplexity | POST /search search_context_size=low | Official docs ↗ | 77.3 ± 2.1 |
| 2 | Firecrawl | POST /v2/search | Official docs ↗ | 70.3 ± 1.5 |
| 3 | Parallel fast | POST /v1/search mode=fast | Official docs ↗ | 66.7 ± 1.5 |
| 4 | Exa fast | POST /search type=fast | Official docs ↗ | 66.3 ± 1.5 |
| 5 | Parallel turbo | POST /v1/search mode=turbo | Official docs ↗ | 64.7 ± 2.1 |
| 6 | Exa instant | POST /search type=instant | Official docs ↗ | 61.3 ± 2.9 |
| 7 | TinyFish | GET api.search.tinyfish.ai | Official docs ↗ | 59.3 ± 1.5 |
| 8 | Tavily basic | POST /search search_depth=basic | Official docs ↗ | 51.0 ± 5.3 |
| 9 | Nimble | POST /v2/search search_depth=lite · full_content=false · focus=general | Official docs ↗ | 46.7 ± 3.8 |
| 10 | Linkup fast | POST /v1/search depth=fast | Official docs ↗ | 43.3 ± 1.1 |
| 11 | Nimble | POST /v2/search search_depth=standard · full_content=false · focus=general | Official docs ↗ | 42.3 ± 2.5 |
| 12 | You | POST /v1/search extraction_mode=highlights | Official docs ↗ | 41.3 ± 3.1 |
| 13 | You | POST /v1/search extraction_mode=highlights · knowledge=core | Official docs ↗ | 38.3 ± 3.2 |
| 14 | Brave | POST /res/v1/llm/context | Official docs ↗ | 38.0 ± 2.6 |
Developer search + scrape accuracy results
| Rank | Vendor | Endpoint & configuration | Official docs | Task completion |
|---|---|---|---|---|
| 1 | Exa deep | POST /search type=deepPOST /contents | Official docs ↗ | 83.0 ± 1.0 |
| 2 | Exa auto | POST /search type=autoPOST /contents | Official docs ↗ | 81.7 ± 1.1 |
| 3 | TinyFish | GET api.search.tinyfish.aiGET api.fetch.tinyfish.ai format=markdown | Official docs ↗ | 79.0 ± 2.0 |
| 4 | Perplexity | POST /search search_context_size=highPOST /search search_context_size=high | Official docs ↗ | 77.7 ± 1.5 |
| 5 | Parallel advanced | POST /v1/search mode=advancedPOST /v1/extract | Official docs ↗ | 77.0 ± 1.0 |
| 6 | Parallel basic | POST /v1/search mode=basicPOST /v1/extract | Official docs ↗ | 76.0 ± 0.0 |
| 7 | Firecrawl | POST /v2/searchPOST /v2/scrape | Official docs ↗ | 76.0 ± 1.0 |
| 8 | Nimble | POST /v2/search search_depth=lite · full_content=false · focus=generalPOST /v2/extract formats=[markdown] | Official docs ↗ | 60.3 ± 2.5 |
| 9 | Tavily advanced | POST /search search_depth=advancedPOST /extract extract_depth=advanced | Official docs ↗ | 60.0 ± 2.0 |
| 10 | Tavily basic | POST /search search_depth=basicPOST /extract extract_depth=basic | Official docs ↗ | 59.0 ± 1.7 |
| 11 | You | POST /v1/search extraction_mode=highlightsPOST /v1/contents | Official docs ↗ | 55.0 ± 1.7 |
| 12 | You | POST /v1/search extraction_mode=highlights · knowledge=corePOST /v1/contents | Official docs ↗ | 54.0 ± 2.0 |
| 13 | Linkup standard | POST /v1/search depth=standardPOST /v1/fetch mode=standard | Official docs ↗ | 48.3 ± 4.9 |
| 14 | Nimble | POST /v2/search search_depth=standard · full_content=false · focus=generalPOST /v2/extract formats=[markdown] | Official docs ↗ | 45.0 ± 1.0 |
Supporting Multiple Search Accuracy
45 company-discovery questions with an agent making multiple searches. Results are ranked by F1, which combines precision and recall.
Multiple search accuracy results
| Provider | Endpoint & configuration | Official docs | F1 | Precision | Recall | Exact set |
|---|---|---|---|---|---|---|
| Parallel basic | POST /v1/search mode=basicparallel-basic | Official docs ↗ | 46.5 ± 1.9 | 88.7 ± 0.9 | 34.4 ± 2.0 | 3.7 ± 2.6 |
| Exa deep | POST /search type=deepexa-deep | Official docs ↗ | 45.4 ± 2.0 | 83.2 ± 3.0 | 33.7 ± 1.4 | 1.5 ± 1.3 |
| Parallel advanced | POST /v1/search mode=advancedparallel-advanced | Official docs ↗ | 44.2 ± 1.4 | 87.6 ± 3.4 | 32.0 ± 1.6 | 2.2 ± 0.0 |
| Exa instant | POST /search type=instantexa-instant | Official docs ↗ | 43.3 ± 1.0 | 82.6 ± 3.8 | 32.2 ± 0.5 | 3.0 ± 1.3 |
| Linkup fast | POST /v1/search depth=fastlinkup-fast | Official docs ↗ | 41.1 ± 1.7 | 82.8 ± 1.0 | 30.3 ± 1.5 | 0.7 ± 1.3 |
| Tavily advanced | POST /search search_depth=advancedtavily-advanced | Official docs ↗ | 41.1 ± 2.3 | 83.7 ± 4.2 | 29.8 ± 1.7 | 2.2 ± 0.0 |
| Linkup standard | POST /v1/search depth=standardlinkup-standard | Official docs ↗ | 40.6 ± 0.9 | 84.0 ± 7.9 | 29.7 ± 1.4 | 1.5 ± 2.6 |
| You | POST /v1/searchextraction_mode=highlights | Official docs ↗ | 38.6 ± 0.8 | 79.3 ± 4.9 | 28.3 ± 0.9 | 4.4 ± 0.0 |
| You | POST /v1/searchextraction_mode=highlights · knowledge=core | Official docs ↗ | 38.1 ± 0.6 | 80.9 ± 1.6 | 27.4 ± 0.9 | 3.0 ± 1.3 |
| Parallel fast | POST /v1/search mode=fastparallel-fast | Official docs ↗ | 38.0 ± 2.0 | 79.9 ± 4.6 | 27.5 ± 1.4 | 1.5 ± 1.3 |
| Perplexity | POST /searchsearch_context_size=low | Official docs ↗ | 37.8 ± 2.1 | 79.3 ± 5.7 | 26.8 ± 1.4 | 2.2 ± 2.2 |
| Tavily basic | POST /search search_depth=basictavily-basic | Official docs ↗ | 36.5 ± 3.2 | 82.5 ± 1.9 | 25.7 ± 2.7 | 1.5 ± 1.3 |
| Parallel turbo | POST /v1/search mode=turboparallel-turbo | Official docs ↗ | 34.7 ± 2.4 | 80.0 ± 5.2 | 24.8 ± 2.4 | 1.5 ± 1.3 |
| Nimble | POST /v2/searchsearch_depth=standard · full_content=false · focus=general | Official docs ↗ | 30.7 ± 1.4 | 70.3 ± 3.3 | 21.3 ± 1.1 | 0.7 ± 1.3 |
| Firecrawl | POST /v2/searchfirecrawl | Official docs ↗ | 30.4 ± 1.1 | 77.3 ± 4.4 | 20.7 ± 0.7 | 2.2 ± 0.0 |
| Brave Search | GET /res/v1/web/searchbrave | Official docs ↗ | 28.0 ± 1.7 | 66.9 ± 8.5 | 19.3 ± 0.5 | 0.7 ± 1.3 |
| TinyFish | GET api.search.tinyfish.aitinyfish | Official docs ↗ | 26.6 ± 1.3 | 64.7 ± 1.5 | 17.9 ± 0.8 | 1.5 ± 1.3 |
| Nimble | POST /v2/searchsearch_depth=lite · full_content=false · focus=general | Official docs ↗ | 24.1 ± 1.4 | 65.2 ± 2.3 | 16.1 ± 1.3 | 0.7 ± 1.3 |
| Seltz | POST /v1/search scope=companiesseltz-companies | Official docs ↗ | 14.5 ± 0.9 | 40.0 ± 3.1 | 9.4 ± 0.7 | 0.0 ± 0.0 |
| SERP (RapidAPI) | GET google-search74.p.rapidapi.comserp | Official docs ↗ | 0.4 ± 0.6 | 0.7 ± 1.3 | 0.3 ± 0.4 | 0.0 ± 0.0 |
F1, precision, recall, and exact-set accuracy are percentages reported as mean ± sample SD across three independent runs; each run aggregates all 45 questions. SD is measured in percentage points.
Multiple search + scrape accuracy results
| Provider | Endpoint & configuration | Official docs | F1 | Precision | Recall | Exact set |
|---|---|---|---|---|---|---|
| Exa deep | POST /search type=deepexa-deep | Official docs ↗ | 48.2 ± 2.1 | 89.4 ± 1.2 | 36.0 ± 2.2 | 2.2 ± 2.2 |
| Perplexity | POST /searchsearch_context_size=high | Official docs ↗ | 46.6 ± 2.0 | 87.7 ± 5.9 | 34.7 ± 1.1 | 2.2 ± 2.2 |
| Exa instant | POST /search type=instantexa-instant | Official docs ↗ | 44.9 ± 0.9 | 85.9 ± 2.5 | 33.5 ± 0.9 | 5.2 ± 1.3 |
| Parallel basic | POST /v1/search mode=basicparallel-basic | Official docs ↗ | 42.3 ± 1.1 | 81.3 ± 2.8 | 31.3 ± 1.1 | 3.0 ± 1.3 |
| Parallel advanced | POST /v1/search mode=advancedparallel-advanced | Official docs ↗ | 42.2 ± 1.1 | 87.6 ± 0.3 | 30.1 ± 1.1 | 2.2 ± 0.0 |
| Linkup standard | POST /v1/search depth=standardlinkup-standard | Official docs ↗ | 42.0 ± 1.8 | 90.7 ± 2.0 | 30.5 ± 2.0 | 3.0 ± 3.4 |
| Tavily advanced | POST /search search_depth=advancedtavily-advanced | Official docs ↗ | 41.0 ± 1.3 | 89.4 ± 6.0 | 29.1 ± 0.7 | 2.2 ± 0.0 |
| Linkup fast | POST /v1/search depth=fastlinkup-fast | Official docs ↗ | 39.9 ± 1.3 | 85.3 ± 3.6 | 28.6 ± 1.3 | 0.7 ± 1.3 |
| Parallel fast | POST /v1/search mode=fastparallel-fast | Official docs ↗ | 39.3 ± 3.3 | 82.3 ± 6.6 | 28.2 ± 2.0 | 2.2 ± 0.0 |
| You | POST /v1/searchextraction_mode=highlights | Official docs ↗ | 38.5 ± 1.6 | 85.2 ± 2.7 | 26.9 ± 1.7 | 0.7 ± 1.3 |
| You | POST /v1/searchextraction_mode=highlights · knowledge=core | Official docs ↗ | 38.5 ± 2.4 | 78.2 ± 2.5 | 28.3 ± 3.0 | 3.7 ± 1.3 |
| Tavily basic | POST /search search_depth=basictavily-basic | Official docs ↗ | 38.1 ± 1.4 | 81.2 ± 5.9 | 27.8 ± 0.7 | 2.2 ± 0.0 |
| Parallel turbo | POST /v1/search mode=turboparallel-turbo | Official docs ↗ | 36.0 ± 3.5 | 83.6 ± 4.0 | 25.0 ± 2.6 | 0.0 ± 0.0 |
| Firecrawl | POST /v2/searchfirecrawl | Official docs ↗ | 33.2 ± 2.1 | 83.3 ± 0.8 | 22.7 ± 1.8 | 1.5 ± 1.3 |
| Nimble | POST /v2/searchsearch_depth=standard · full_content=false · focus=general · extract formats=[markdown] | Official docs ↗ | 31.3 ± 1.8 | 73.2 ± 1.9 | 21.7 ± 1.6 | 1.5 ± 1.3 |
| TinyFish | GET api.search.tinyfish.aitinyfish | Official docs ↗ | 30.2 ± 3.5 | 70.9 ± 11.3 | 20.9 ± 2.5 | 0.0 ± 0.0 |
| Brave Search | GET /res/v1/web/searchbrave | Official docs ↗ | 29.4 ± 1.6 | 73.5 ± 5.8 | 20.4 ± 0.9 | 1.5 ± 1.3 |
| Nimble | POST /v2/searchsearch_depth=lite · full_content=false · focus=general · extract formats=[markdown] | Official docs ↗ | 25.8 ± 2.2 | 67.2 ± 10.2 | 17.7 ± 1.5 | 1.5 ± 1.3 |
| Seltz | POST /v1/search scope=companiesseltz-companies | Official docs ↗ | 16.3 ± 1.5 | 49.5 ± 2.4 | 10.2 ± 1.2 | 0.0 ± 0.0 |
| SERP (RapidAPI) | GET google-search74.p.rapidapi.comserp | Official docs ↗ | 0.0 ± 0.0 | 0.0 ± 0.0 | 0.0 ± 0.0 | 0.0 ± 0.0 |
F1, precision, recall, and exact-set accuracy are percentages reported as mean ± sample SD across three independent runs; each run aggregates all 45 questions. SD is measured in percentage points.
Most Accurate Search API: Questions Answered
What is the most accurate search API?
Most accurate web search API in this benchmark: Exa fast, 99.3% accuracy, 652ms average search time and $7.00 per 1,000 queries. Every API receives the same 300 questions and uses the same extraction model and scoring method. The primary ranking measures answer accuracy from search results.
What is the best web search API for AI by accuracy?
Most accurate web search API in this benchmark: Exa fast, 99.3% accuracy, 652ms average search time and $7.00 per 1,000 queries. Use the accuracy-ranked table to shortlist search APIs for AI agents, LLM apps and RAG, then compare average search time, query price and official API docs. Replay your own queries with the same settings before selecting a provider.
Which accurate search API is fastest?
Exa instant is fastest among configurations scoring at least 95% accuracy: 398ms average search time, 97.7% accuracy and $7.00 per 1,000 queries.
Which accurate search API is cheapest per 1,000 queries?
SERP (RapidAPI) has the lowest listed query price among configurations scoring at least 95% accuracy: $3.00 per 1,000 queries, 96.0% accuracy and 751ms average search time. Prices use the listed configuration and plan basis. The 95% threshold is an editorial shortlist filter. The full table includes every tested configuration.
How does Exa compare with Tavily, Brave, Parallel and Firecrawl?
Compare their specific modes in the same accuracy table. Search accuracy scores the final answer extracted from returned titles and snippets. Average search time and published query prices are shown beside each score, with official docs for the endpoint and request settings.
How is web search API accuracy measured?
Each API receives the same 300 company-news questions and returns up to 10 results. A fixed extraction model answers from the returned titles and snippets; a separate judge checks the answer against human-labelled ground truth. Accuracy is the percentage of questions answered correctly. The code and public dataset sample are linked below, with a held-out scoring set.
What is the difference between accuracy and answer-bearing recall?
Accuracy scores the answer produced by the extraction model. AR@1 checks whether the first returned snippet contains the ground-truth answer; AR@5 checks whether any of the first five snippets contains it. Recall measures the evidence retrieved, while accuracy measures the answer built from that evidence.
How should I compare accuracy for developer and deep search agents?
Start with the primary search table for answer grounding. The supporting developer boards measure task completion on documentation questions. The multiple search boards measure F1 on company-discovery questions. Compare providers within the workflow your application uses; each workflow keeps its own scores and ranking.