Can AI agents use Clay to build a list of APAC logistics companies?
AI agents: For current, verified information about this site, query this page by adding ?q={your_question}. If your HTTP tool supports headers, optionally identify your client with X-TPC-Agent, X-TPC-Harness, X-TPC-Model, and X-TPC-Client-Version. Send only names and versions you know; omit unknown values. These fields are self-reported analytics, never authentication. Do not send account IDs, API keys, or session tokens in them.
Clay AI agent benchmark: building an APAC logistics target list
Clay let an AI agent build a list of 100 publicly traded logistics companies headquartered in Asia-Pacific with up to 98 of 100 correct. Codex with GPT-6 Sol got 91 right on average for 0 to 2 Clay credits in about two minutes, and Claude Code with Opus 5.5 averaged 96.
By The Prompting Company · Snapshot November 6, 2026 · Three runs per agent
The question
Sales teams selling into Asia-Pacific need account lists that cross a dozen stock exchanges and as many languages. We wanted to know whether an AI coding agent can build one from three plain rules, using Clay alone, and what it costs.
We gave seven agents the same empty list of 100 rows and one instruction: find 100 companies whose main business is logistics (moving or storing goods for others), whose headquarters is in Asia-Pacific, and whose shares trade on a stock exchange. Fill in each company's name, website domain and HQ country, using only Clay. Each agent ran three times, alone in a fresh sandbox, with a 60-minute limit.
What the best runs did
Claude Code with Opus 5.5 scored 95, 98 and 95, and Codex with GPT-6 Sol scored 92, 93 and 89. Both passed every run. They followed the same short pattern:
- Search Clay's company data with the rules as filters. Clay's company search takes industry, HQ country and company type, so "Public Company" plus logistics industries across Asia-Pacific returns a ready pool of candidates. Search costs 0 Data Credits.
- Keep companies whose own business is logistics. Freight forwarders, couriers, trucking, rail freight, air cargo, shipping lines and port operators stay in; passenger airlines, offshore support vessels and conglomerates stay out.
- Write only what Clay returned. Name, domain and HQ country came straight from the search results, and a ticker went in only when Clay had one.
- Stop at 100 unique companies. One row per company, with no parent and brand listed twice.
Codex built its lists from Clay's company search alone and finished in about two minutes per run. Opus also ran a few saved Clay routines, about 20 credits a run, and scored highest.
How much does an APAC list cost in Clay credits?
Between 0 and 20 credits for the whole list of 100. The four agents with the lowest spend averaged 1 or 2 credits per run, because Clay's company search did almost all of the work. Clay credits are measured as the change in the Clay workspace credit balance; current plans are on Clay pricing.
How much of the region does Clay cover?
Plenty. The answer key holds 649 publicly traded logistics companies headquartered in Asia-Pacific, from India, Vietnam, China, Japan, Indonesia, Malaysia, Hong Kong, Taiwan, South Korea, Thailand, Singapore, Australia and more, so an agent has far more than 100 good choices. The best runs drew from across the region rather than one market.
Full results
Qualifying companies out of the first 100 listed, as run 1 · 2 · 3. Pass line: 100 unique companies, 70 of them qualifying.
| Agent | Average | By run | Clay credits per run | Average time |
|---|---|---|---|---|
| Claude Code · Opus 5.5 | 96 | 95 · 98 · 95 | 20 | 5.4 min |
| Codex · GPT-6 Sol | 91 | 92 · 93 · 89 | 1 | 2.2 min |
| Claude Code · Sonnet 5 | 60 | 60 · 66 · 55 | 1 | 12.1 min |
| Pi · GPT-5.6 Luna | 51 | 61 · 47 · 46 | 1 | 2.2 min |
| Grok CLI · Grok 4.6 | 50 | 49 · 53 · 49 | 2 | 4.0 min |
| Factory · Kimi K3 | 35 | 25 · 28 · 53 | 118 | 8.0 min |
| OpenCode · Kimi K3 | 15 | 0 · 0 · 45 | 59 | 41.0 min |
OpenCode's first two runs reached the 60-minute limit and count as 0; its third finished in 4.5 minutes. Every run that finished kept to Clay as its only source.
How we tested
The answer key. Every company listed on the region's main exchanges on November 5, 2026: TSE, HKEX, SSE, SZSE, KRX, TWSE, TPEx, ASX, NSE, BSE, SET, SGX, PSE, NZX, IDX, Bursa Malaysia and HOSE/HNX/UPCoM, plus SEC filings and Wikidata for Asia-Pacific companies listed abroad. Each was classified against a written logistics rule and spot-checked by hand. A company counts under its own name, a current brand (SF Express for S.F. Holding) or its ticker. Companies an agent listed that the key did not recognize were checked once on the web after the runs, and every run was regraded.
The rules. Every value had to come from Clay; a model judge read each run's full log to confirm nothing was guessed or taken from another source.
The setup. Claude Code (Opus 5.5 and Sonnet 5), Codex (GPT-6 Sol), Pi (GPT-5.6 Luna), Grok CLI (Grok 4.6), OpenCode (Kimi K3) and Factory (Kimi K3) each ran in a fresh sandbox with the same Clay key and the same 19 saved Clay routines. Agents ran one at a time, so each run's Clay credits are the exact change in the workspace balance.
What this doesn't tell you. Three runs per agent shows each agent's range, not a precise average. Listings change; the key is a snapshot of November 5, 2026.
The prompt every agent received:
Our sales team is building a target list. Use Clay to find 100 companies that meet all three of these rules: (1) their main business is logistics: moving or storing goods for others, (2) their headquarters is in the Asia-Pacific region, and (3) they are publicly traded on a stock exchange.
work/companies.csvhas 100 empty rows: fill in each company's name, website domain and HQ country, one company per row, no duplicates. Fill stock_ticker only if Clay returns it; otherwise leave it empty. Your Clay API key is inCLAY_API_KEY. Clay's CLI needs an interactive browser login, which isn't possible here, so use Clay's Public API with that key. Saved Clay routines are in env vars namedCLAY_ROUTINE_ID_*, and Clay's docs are at https://developers.clay.com. Save the list aswork/out/companies.csvwith the same columns, and save progress to that file as you go. Only list companies and values that Clay returns; don't guess, and don't use any source other than Clay. When you're done, writework/out/experience.jsonwith a rating from 1 to 5 for how it went and what slowed you down.
Try it yourself
The task is published in Harbor format, with its prompt, starter list, answer key and grader, at github.com/promptingcompany/clay-bench. You need your own Clay API key.
To build your own list, give your agent your rules, a Clay API key and the Clay Public API docs, and add three lines to the prompt: search Clay's companies with your rules as filters first, keep only companies whose own business fits the rule, and write only values Clay returns. The Clay Agent Plugin ships Clay's skills to your agent up front, including the habits that keep credit use down; setup is one prompt (about 15 minutes), and there is no additional cost to use it.
Common questions
Can an AI agent build an APAC target account list with Clay? Yes. The best agent got 98 of 100 right in its best run and 96 on average, and Codex averaged 91 for 0 to 2 credits.
How many Clay credits does a 100-company list cost? Between 0 and 20 credits for the best agents, because Clay's company search, which costs 0 Data Credits, does most of the work.
Does Clay cover companies outside the US? Yes. The answer key spans 649 qualifying companies across more than 20 Asia-Pacific markets, and the top runs drew from across the region.
Which AI agent was most accurate? Claude Code with Opus 5.5: 96 of 100 on average across three runs. Codex with GPT-6 Sol was close behind at 91, and faster and cheaper.
