Parallel

Agent-optimized search index T2 Search /parallel

Parallel is a web search API designed from the ground up for AI agents rather than human clickers. Agents express intent as a natural-language objective and Parallel returns the most relevant sources as token-dense, compressed excerpts — pre-trimmed to fit a context window, which means higher accuracy, fewer round-trips and lower token spend. That agent-first design shows in the numbers, with category-leading accuracy on demanding research benchmarks, and its search sits alongside a broader suite for deep research, extraction and continuous monitoring.

Entry price
$5 (base) / $9 (pro) / 1k
MCP
official + no-key MCP
Latency
P50 <1s

Call Parallel through the gateway

Same paths, parameters and responses as Parallel's native API — auxiliar injects the upstream key server-side and debits the provider's metered cost from your credits.

curl https://api.auxiliar.ai/parallel/<native-path> \
  -H "Authorization: Bearer $AUXILIAR_API_KEY"

Exact endpoints and parameters: docs.auxiliar.ai/parallel. An empty balance returns HTTP 402 — no silent degradation; see how billing behaves.

Choose if

You want the lowest-friction in-loop option — a no-key Search MCP and sub-second P50.

Avoid if

You need a relevance score, full content or a synthesized answer in the Search response.

Evals

How Parallel performed on the curator-fleet web-access evals — the real measured numbers behind each composite, ranked best-first. Every provider runs the identical task corpus; each scorecard carries its run date. Latest run: 2026-06-21.

Answercomposite 7.58/10 #3 of 7 · 2026-06-21
1.00
Correctness
0.48
Citation faithfulness
16.3 s
Latency p50
$0.01 / call
Cost
0%
Errors
Searchcomposite 6.76/10 #5 of 11 · 2026-06-20
0.64
Recall@10
2.7 s
Latency p50
$0.0062 / useful result
Cost
0%
Errors
← All tools