Web-grounded Q&A with citations
Answers must reflect live web with citations without DIY search stack.
At a glance — Oct 2026
Updated 2026-10-27 · forward this page: /jtbds/web-grounded-qa#briefScores as of 2026-09-14 (43 days old; Depot marks scores stale after 30 days). The pick is still Depot's current editorial pick; it has not been re-reviewed since. Prices below carry their own dates.
- Depot Pick
- Perplexity · 7.67/10 for this job (curated estimate, high confidence)
capability groundedness + dx; lock-in medium.
- Runner-up
- Exa · 8.30/10
Scores higher than the pick on: Capability fit, Price / performance, Latency / reliability, Developer & agent DX, Privacy / data retention, Lock-in risk (higher = worse).
- When not to use the pick
- Need full control of retrieval/ranking — compose Exa/Tavily + your LLM.
- Lock-in / migration
- Runner-up Exa has lower stored lock-in risk (4.5 vs 5.5, 0–10, higher = more lock-in).
- Sourced price (no comparable unit)
Pick's lowest sourced SKU: pplx-embed-v1-0.6b $0.004 (usd_per_million_tokens) — source, as of 2026-10-03, re-checked 2026-10-27
Cheapest matching SKU: Perplexity pplx-embed-v1-0.6b $0.004 — source, as of 2026-10-03, re-checked 2026-10-27. The pick is the cheapest.
Price history: first seen 2026-10-03 · last changed no change recorded · history JSON
- Free option
- Gemini 2.5 Flash Google Search grounding free requests (free) — source, as of 2026-10-26
- Failure modes to watch
- Prompt injection: Compare each tool call to the original user task. If the call only makes sense given text that arrived from a fetched page, file, or message, stop and ask. Record which source the instruction came from. This catalog does not store payloads or attack steps.
- Runaway spend: Set a budget and a tool-call cap per task. Alert when spend jumps. Stop the run at the cap instead of summarizing a partial trace as success.
- Stale retrieval: Store the source URL and the time it was fetched. If a newer canonical page exists, or two stored sources disagree, do not answer from the older one.
- Ungrounded answer: Require a source URL or passage id on factual sentences. Drop or mark any sentence whose source does not contain the claim.
- Context overflow: If the input exceeds the window, say so and retrieve a narrower slice. Do not answer from a truncated head as if it were the whole source.
- Instruction drift: Restate the stop condition in the trace. End the run when it is met. Review long traces for steps that do not serve the original task.
Cite: Agent Provider Depot, Depot Pick for web-grounded-qa: perplexity, scores as of 2026-09-14. Agents: GET /api/decide?jtbd=web-grounded-qa
Free Web-grounded Q&A with citations options · for research agents
Perplexity
capability groundedness + dx; lock-in medium.
Pricing (directional): As of 2026-09-14 (docs.perplexity.ai pricing): Sonar $1/$1 per 1M in/out; Sonar Pro $3/$15; Sonar Reasoning Pro $2/$8; Sonar Deep Research $2/$8 + citation $2/1M + search queries $5/1K + reasoning $3/1M. Request fees per 1K (low/med/high co…
Lock-in risk: higher = more lock-in (worse). Overall already inverts this dimension.
Use when you want grounded answers without building search+cite yourself. For full agent stacks, prefer Exa/Tavily + your LLM.
Quote for Web-grounded Q&A with citations
Depot Pick stays Perplexity. Prices below are sourced SKU rows only (as of 2026-10-27). Input and output are not blended. GET /api/quote?jtbd=web-grounded-qa
Pick is the cheapest sourced input $/1M among candidates (0.004).
- pplx-embed-v1-0.6b usd_per_million_tokens $0.004 source as of 2026-10-03
- pplx-embed-v1-4b usd_per_million_tokens $0.03 source as of 2026-10-03
- pplx-embed-context-v1-0.6b usd_per_million_tokens $0.008 source as of 2026-10-03
- pplx-embed-context-v1-4b usd_per_million_tokens $0.05 source as of 2026-10-03
- web search tool search_type=web usd_per_1k_calls $2.5 source as of 2026-10-03
2 of 8 candidates have no sourced price row. 5 are excluded from the Input $/1M comparison (chat_completion_input). min sourced input SKU among candidates; not quality-adjusted
Runners-up
Exa
Depot pick for research-style agent search. Pair with Firecrawl when you need full page content.
Tavily
Depot co-default for agent web search with Exa. Pick Tavily for classic SERP-like agent search.
OpenAI
Default when you want maximum capability and ecosystem glue. Multi-home critical paths behind OpenAI-compat…
You.com API
Agent search peer with Answer/Research SKUs. Exa remains depot agent-search pick; You.com strong when ZDR +…
When NOT to use the top pick
Need full control of retrieval/ranking — compose Exa/Tavily + your LLM.
Full ranking (score table)
Ranked with this JTBD's weights. Scores are curated estimates unless marked otherwise. Agents: prefer /api/decide then this table for tie-breaks.
| # | Provider | Overall | Fit | $/perf | Latency | DX | As of |
|---|---|---|---|---|---|---|---|
| 1 | Exa | 8.3 | 9.0 | 8.2 | 8.4 | 9.2 | 2026-10-19 |
| 2 | Tavily | 8.3 | 8.8 | 8.5 | 8.4 | 9.2 | 2026-09-14 |
| 3 | Linkup | 8.2 | 8.4 | 8.6 | 8.2 | 8.8 | 2026-09-14 |
| 4 | OpenAI | 8.2 | 9.6 | 6.2 | 8.8 | 9.5 | 2026-09-14 |
| 5 | Parallel | 8.1 | 8.8 | 8.4 | 8.0 | 8.8 | 2026-09-14 |
| 6 | You.com API | 8.1 | 8.4 | 7.8 | 8.2 | 8.8 | 2026-09-14 |
| 7 | xAI | 7.7 | 8.8 | 8.0 | 7.8 | 7.8 | 2026-09-14 |
| 8 | PerplexityDepot Pick | 7.7 | 8.2 | 7.5 | 8.0 | 8.0 | 2026-09-14 |
Score methodology for this job
capability groundedness + dx; lock-in medium.
capabilityFit: 25%pricePerformance: 18%latencyReliability: 15%developerDx: 15%privacyPosture: 10%lockInRisk: 7%maturity: 10%