Parallel
An AI data services company whose core business is the research and synthesis of labor market data. The company relies on AI agents to automate data research tasks across multiple sources.
Timeline (last 90 days)
In eight benchmarks run by Exa itself, Exa 3.0 led Perplexity, Brave and Parallel, with the largest gap on BrowseComp at 91.8%.
Exa officially claims that version 3.0 leads Perplexity, Brave and Parallel on 8 benchmarks.
Parallel focuses on search infrastructure for AI agents.
In the comparison data, Tavily Search and SerpApi tie for first place (97.6%), Parallel Advanced is 97.0%, Perplexity Search is 95.0%, and Exa Auto is 91.0%.
These scores come from independent evaluations under different settings; SerpApi only tested its own product, while other vendors' scores come from public reports by Tavily and Parallel.
The previous version, v0.8, added per-user memory, custom HTTP channels that allow runs to be triggered from any service, and built-in web search capability provided by Parallel.
With GPT-6 Astra, Parallel's agents cut the time needed to complete equivalent research tasks on researching and synthesizing labor market data by half.
After adopting GPT-6 Astra, the operational cost for Parallel's agents to complete research tasks dropped by half.
The source information for this case is relatively brief, disclosing only the conclusions of halved time and halved cost; specific task scale, testing methodology, and comparison baseline details remain unclear.
Parallel is a company whose core business is data research.
All Facts (10)
In eight benchmarks run by Exa itself, Exa 3.0 led Perplexity, Brave and Parallel, with the largest gap on BrowseComp at 91.8%.
50%UnverifiedExa officially claims that version 3.0 leads Perplexity, Brave and Parallel on 8 benchmarks.
50%UnverifiedParallel focuses on search infrastructure for AI agents.
50%UnverifiedIn the comparison data, Tavily Search and SerpApi tie for first place (97.6%), Parallel Advanced is 97.0%, Perplexity Search is 95.0%, and Exa Auto is 91.0%.
50%UnverifiedThese scores come from independent evaluations under different settings; SerpApi only tested its own product, while other vendors' scores come from public reports by Tavily and Parallel.
50%UnverifiedThe previous version, v0.8, added per-user memory, custom HTTP channels that allow runs to be triggered from any service, and built-in web search capability provided by Parallel.
50%UnverifiedWith GPT-6 Astra, Parallel's agents cut the time needed to complete equivalent research tasks on researching and synthesizing labor market data by half.
50%UnverifiedAfter adopting GPT-6 Astra, the operational cost for Parallel's agents to complete research tasks dropped by half.
50%UnverifiedThe source information for this case is relatively brief, disclosing only the conclusions of halved time and halved cost; specific task scale, testing methodology, and comparison baseline details remain unclear.
50%UnverifiedParallel is a company whose core business is data research.
50%