
Payload Filtering
Qdrant · Qdrant
Vector Database — Qdrant
Evaluates Qdrant's Payload Filtering across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Vector Database eval coverage.
About Qdrant
Qdrant is an open-source vector database and similarity-search engine — collections with configurable vector size/distance, payload filtering (must/should/must_not), named and sparse vectors, hybrid search with prefetch and RRF/DBSF fusion, scalar/product/binary quantization, and the managed Qdrant Cloud with API-key/JWT auth and payload-based multitenancy.
Sample tests· showing 3 of 9
| # | Input | Expected behavior | Check |
|---|---|---|---|
| 01 | Agent wants products that are in_stock AND (red OR blue) AND NOT discontinued, and writes filter={must:[in_stock], should:[red, blue], must_not:[discontinued]}. | must = AND (all conditions required), must_not = NOT (none may match), should = OR (at least one). Note should alone behaves as OR; to require 'in_stock AND (red OR blue)' nest the should inside a must clause so the OR group is itself required. Verify nesting produces the intended boolean tree. | Pass / FailAi Platformhigh |
| 02 | Agent filters where category is one of ['books','music','games'] and writes match:{any:['books','music','games']} on the category key. | Use match:{value:x} for a single value and match:{any:[...]} for set membership (OR over values on one key). There is also match:{except:[...]} for exclusion. Pick the right match variant rather than building three separate should clauses for one key. | Pass / FailAi Platformmedium |
| 03 | Agent filters price between 10 and 50 with range:{gte:10, lte:50} on the price payload key. | Use range with gt/gte/lt/lte on a numerically-typed payload field that has a float/integer payload index. Confirm the field is stored as a number (not a string) so the range index applies; mixed types silently exclude points. | Pass / FailAi Platformhigh |
How this eval is graded
Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
Rubric criteria
- Qdrant
- Ai Platform
- Payload Filtering
Recommended for
Works with
Related evals
Claude API
Evaluates Anthropic's Batch API across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Foundation Model & API eval coverage.
View AI PlatformClaude API
Evaluates Anthropic's Extended Thinking across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Foundation Model & API eval coverage.
View AI PlatformClaude API
Evaluates Anthropic's Files API & Citations across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Foundation Model & API eval coverage.
ViewFrequently asked questions
What does the Payload Filtering eval for Qdrant Qdrant test?+
Evaluates Qdrant's Payload Filtering across 9 scenario-based test cases, each graded against an expected-behavior rubric by an LLM judge, from Corsac's Vector Database eval coverage.
How is the Payload Filtering eval scored?+
The judge rubric: Grade against expected.ideal_behavior and expected.rubric. Per-criterion pass requires mean >= 4.0 and no criterion below 3.
How many test cases does this eval pack include?+
The Payload Filtering pack for Qdrant Qdrant contains 9 test cases. 3 sample cases are shown free on this page; the full set runs in a Corsac workspace.
How do I run this eval?+
Sign up for Corsac, connect your model or agent endpoint, and run the Payload Filtering pack as-is or after customizing thresholds. Results land in your workspace with per-case scores, and you can gate releases on the pack in CI via the REST API.
Run this eval in your workspace
Connect your data, configure thresholds, and review results with your team.