A1 ratingsdeepsetHaystack

Haystack

deepset · framework · Any industry
B
BGrade
56Score
±1195% range
#322Rank of 1134
0%From fleets
FR-CAT-00916Catalog no.

The six pillars

Each bar is the score; the bracket is its 95% range.

Can you see what it costs?

Only through a gateway you run · prices published
57±27
Desk research 60 · counts for 25% · A cost API you can call scores highest; an admin console or cloud billing export next; a card statement or emailed invoice is all some products leave you. Published prices add to it.

Can you see how much it's used?

No usage API found
50±27
Desk research 50 · counts for 15% · A usage API (tokens, seats, credits, conversations) a buyer can read.

Can you tie cost to one agent?

Exact, per key or project
93±13
Desk research 100 · counts for 15% · Exact (one key or project per agent) beats by-project, by-workspace or by-user.

Can an agent on it report out?

SDK · Self-hosted; add tracing/callbacks or route LLM calls via a gateway.
64±27
Desk research 70 · counts for 15% · A remote MCP connector scores highest, then a local MCP server, an SDK, an API or a workflow step; copy-and-paste lowest. Any open standard counts, not just FleetRegistry.

Can you control spend and access?

Spend cap: None · Per-agent keys: One shared key · Audit log: None · Revoke: None · SSO: None
14±27
Desk research 0 · counts for 20% · Hard spend caps (30%), scoped per-agent keys (25%), an audit log you can read by API (20%), revoking by API (15%), SSO with SCIM (10%).

Does it work on small plans?

Small plans don't get the data
14±27
Desk research 0 · counts for 10% · Whether the plan a person or small business buys gets the cost and usage data at all.

Where the evidence comes from

Independent fleets0%Desk research74%Neutral starting point26%

Once three independent fleets report on a pillar, desk research counts for at most a third of it. FleetRegistry's own fleets never count.

Controls on record

Spend capNone
Per-agent keysOne shared key
Audit logNone
RevokeNone
SSONone
PricesPublished

Open-source (Apache-2.0) self-hosted framework with no buyer admin controls of its own; access controls and auditability are features of the paid Haystack Enterprise Platform (deepset), not the framework.

History

11-12 B 5611-13 B 5611-14 B 5611-15 B 5611-16 B 5611-17 B 5611-18 B 5611-19 B 56
First rated 2026-10-02 at 55 (B). Now 56. Published grades are sealed daily into the transparency log (N4.4). Grade held since 2026-10-02.

Every fact, with its rung and source

Each fact stands on its own source and its own rung of the evidence ladder (Rule N2.1). Verified Observed Declared Inferred. The weakest fact in a pillar sets its weight.

FactValueRungCheckedSources
How a buyer sees cost
Cost
Only through a gateway you run Declared
medium confidence
2026-10-23
1 earlier
[1][2] Correct
Prices published
Cost
Yes Declared
medium confidence
2026-10-23
2 earlier
[1][2] Correct
Usage API
Usage
None found Declared
medium confidence
2026-10-23
1 earlier
[1][2] Correct
Cost per agent
Per agent
Exact, per key or project Verified 2026-10-26
2 earlier
[1][2][3][4] Correct
How an agent reports out
Reports out
SDK Declared
medium confidence
2026-10-23
1 earlier
[1][2] Correct
Per-agent keys
Controls
One shared key Declared
medium confidence
2026-10-23
2 earlier
[1][2] Correct
Audit log
Controls
None Declared
medium confidence
2026-10-23
2 earlier
[1][2] Correct
Revoke access
Controls
None Declared
medium confidence
2026-10-23
2 earlier
[1][2] Correct
Spend cap
Controls
None Declared
medium confidence
2026-10-23
2 earlier
[1][2] Correct
SSO
Controls
None Declared
medium confidence
2026-10-23
2 earlier
[1][2] Correct
Small plans get the data
Small plans
No Declared
medium confidence
2026-10-23
1 earlier
[1][2] Correct

Prices · Free, Apache 2.0

Notes · Costs are LLM API, vector DB, and compute that you run. Agents track step_count, token_usage and tool calls for cost control. deepset sells Haystack Enterprise support (pricing by company size) and the Haystack Enterprise Platform; no public prices.

Scored with the rubric for Open-source frameworks (Rule N7.1). Each pillar blends a neutral 50, desk facts weighted by rung, and independent fleets. A1 is 85+ with fleet proof, A 70, B 55, C 40, D 25, E below. Part N of the Rules

Corrections

1 awaiting review. Accepted corrections are published in the corrections register with an FR-C number.

Run agents on Haystack?

Give each one a permanent number, a named person who answers for it, and a record of what it can touch and what it costs. Free.