AP
Agent Provider Depot
Opinionated picks for agents
OpenAPI
← All jobs

Vector + object-storage hybrid memory

Store embeddings cheaply atop object storage while keeping raw blobs (PDFs, images) co-located for agent re-read without double egress bills.

At a glance — Oct 2026

Updated 2026-10-27 · forward this page: /jtbds/vector-object-hybrid-memory#brief

Scores 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
Turbopuffer · 8.35/10 for this job (curated estimate, high confidence)

Cost architecture pick. Confidence: high for Turbopuffer object-storage thesis.

Runner-up
Tigris · 8.30/10

Scores higher than the pick on: Developer & agent DX, Lock-in risk (higher = worse).

When not to use the pick
Managed Pinecone simplicity wins, or R2/Tigris alone if you bring your own ANN.
Lock-in / migration
Runner-up Tigris has lower stored lock-in risk (3.2 vs 4.5, 0–10, higher = more lock-in).
Sourced price (no comparable unit)

Pick: no sourced price in this job's unit (no_sourced_price).

No cheapest SKU (no_comparable_unit_for_job). No price was invented.

Free option
No free resource lists this job. A cheap paid SKU is not a free option.
Failure modes to watch
  • 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.
  • 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.
Reported outcomes (submitted reports only)

This job: loading… JSON · reports never change scores · seed/test records excluded

Depot Pick (Turbopuffer) on this job: loading… JSON · reports never change scores · seed/test records excluded

Cite: Agent Provider Depot, Depot Pick for vector-object-hybrid-memory: turbopuffer, scores as of 2026-09-14. Agents: GET /api/decide?jtbd=vector-object-hybrid-memory

Depot Pick

Turbopuffer

Cost architecture pick. Confidence: high for Turbopuffer object-storage thesis.

Pricing (directional): As of 2026-10-24 (turbopuffer.com/pricing, fetched 2026-10-24): Usage-based with plan minimums: launch '$16/month minimum usage', scale '$256/month minimum usage', enterprise '>= $4,096/month' with '35% usage premium'. 'You'll pay for actua…

Cite: Depot Pick for vector-object-hybrid-memory: turbopuffer
Overall
8.3
curated_estimate
confidence: high · as of 2026-09-14
Capability fit
8.6
Price / performance
9.4
Latency / reliability
8.2
Developer & agent DX
8.5
Privacy / data retention
7.8
Lock-in risk (higher = worse)
4.5
Maturity / track record
7.0

Lock-in risk: higher = more lock-in (worse). Overall already inverts this dimension.

Sources: turbopuffer pricing: https://turbopuffer.com/pricing · turbopuffer pricing changelog: https://turbopuffer.com/docs/pricing-log · Turbopuffer: https://turbopuffer.com · Turbopuffer docs: https://turbopuffer.com/docs

When your vector bill scares you and much of the corpus is cold, Turbopuffer is the interesting answer.

Quote for Vector + object-storage hybrid memory

Depot Pick stays Turbopuffer. Prices below are sourced SKU rows only (as of 2026-10-27). Input and output are not blended. GET /api/quote?jtbd=vector-object-hybrid-memory

Nothing was compared. no comparable sourced unit for this job; chat input $/1M and other units were not borrowed pickIsPriceWinner is null.

Sourced price unavailable for the Depot Pick. No row in the pricing index.

3 of 5 candidates have no sourced price row. 5 are excluded from the Matching price comparison (none). no comparable sourced unit for this job; chat input $/1M and other units were not borrowed

Runners-up

When NOT to use the top pick

Managed Pinecone simplicity wins, or R2/Tigris alone if you bring your own ANN.

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.

#ProviderOverallFit$/perfLatencyDXAs of
1Cloudflare R28.89.09.59.28.82026-10-19
2TurbopufferDepot Pick8.38.69.48.28.52026-09-14
3Tigris8.38.29.28.08.62026-09-14
4Amazon S38.29.87.09.88.02026-09-14
5Pinecone7.59.06.29.08.82026-09-14

Score methodology for this job

Cost architecture pick. Confidence: high for Turbopuffer object-storage thesis.

  • capabilityFit: 18%
  • pricePerformance: 32%
  • latencyReliability: 14%
  • developerDx: 12%
  • privacyPosture: 8%
  • lockInRisk: 8%
  • maturity: 8%