MarkTechPost ↗ · 2026-11-18 15:10 · UTC+8
SID AI Releases SID-2 Agentic Search Model With 0.720 NDCG Swarm Mode
Model updates MarkTechPost Source language: English
SID AI released its second‑generation agentic search model SID‑2 and the 12‑agent SID‑2 Swarm, which achieved a top‑of‑table 0.720 NDCG on the company’s benchmark, surpassing GPT‑6 Astra’s 0.704, with a 262k token context window and API pricing of $0.25 per M input tokens and $1.00 per M output tokens.
Full text
SID AI has released SID-2, its second-generation agentic search model, alongside SID-2 Swarm. SID-2 is an agentic search model from SID AI that searches your data, reads results and returns ranked document IDs. It replaces fixed RAG pipelines with a model that decides what to search next.
TL;DR
Size: Context window is 262k tokens per the report (256k on sid.ai/models).
Runs on: Hosted API at platform.sid.ai, plus Baseten and SageMaker reference paths. SID’s latency figures assume 1 B200 GPU at 8 concurrent queries.
Performance: A single agent matches Claude Opus 5.5 on retrieval quality. Swarms beat every model SID tested.
Best: SID-2 Swarm (12 agents) scores 0.720 NDCG, above GPT-6 Astra at 0.704.
Worst: SID-2 Low scores 0.556 NDCG, 0.148 below GPT-6 Astra. Astra still leads on California law, SEC filings and FiQA.
Bottom line: Best: frontier-level retrieval at a fraction of the latency. Worst: no disclosed size, weights or independent evals yet.
What is SID-2 and how does it work?
SID-2 is a small search specialist, not a general chat model. A static pipeline runs rewrite, embed, search and rerank in a fixed order. SID-2 instead searches, reads results, adapts and searches again.
Developers implement the search tools against their own backend. A text_search tool covers keyword search and a search tool covers vector search. The model ends each episode by calling report_helpful_ids with a ranked list. No reindexing or data movement is required, since SID-2 only sees what your tools return.
SID-2 Swarm runs several agents that message each other and fork their context into new agents. Each agent explores a separate slice of the corpus. SID argues search gains more from reading more documents than from reasoning harder about each one.
What are reasoning effort and Metric Steering?
SID-2 offers low, medium and high reasoning effort. Each level was trained as a separate expert, then merged with multi-teacher on-policy distillation. Metric Steering lets users target NDCG, F1 or Recall@k, with k from 1 to 50. On SID’s 506-question validation set, SID-2 at high effort scores 0.14 to 0.22 higher per metric than the base model.
How was SID-2 trained?
Training used 2-stage reinforcement learning. Stage 1 optimizes retrieval quality only. Stage 2 adds a multiplicative latency penalty that applies only to positive rewards. A proposer-solver self-play loop generated hard queries against a frozen SID-1 solver. SID reports a 6x training-throughput gain over SLIME’s default setup.
How does SID-2 perform on benchmarks?
SID built a 506-question benchmark across 19 corpora. These include HotpotQA, Enron emails, SEC filings, PubMed and BrowseComp-Plus. All numbers below are self-reported from Tables 2 to 4 of the report.
SID-2 Medium: 0.585 NDCG in 3.11 s, versus Claude Opus 5.5 (medium) at 0.577 in 25.1 s.
SID-2 Swarm, 6 agents: 0.701 NDCG in 17.7 s, versus GPT-6 Astra (max) at 0.704 in 135 s.
SID-2 Swarm, 12 agents: 0.720 NDCG, the highest score in the table.
Baselines: BM25 scores 0.260 and vector search plus reranker scores 0.287.
Swarm leads on 11 of 19 corpora. On WixQA, a single high-effort agent matches Astra at about 232x lower cost. At its lowest effort, SID-2 beats 4 parallel SID-1 runs with 13x fewer FLOPs.
SID-2 vs GPT-6 Astra vs Claude Opus 5.5
FeatureSID-2 (medium)SID-2 Swarm (6 agents)GPT-6 Astra (max)Claude Opus 5.5 (medium)
TypeAgentic search modelMulti-agent searchGeneral LLMGeneral LLM
ParametersNot disclosedNot disclosedNot disclosedNot disclosed
Context262k (report) / 256k (site)Effectively infinite (per SID)Not disclosed1M
NDCG, SID benchmark0.5850.7010.7040.577
Inference time per query3.11 s17.7 s135 s25.1 s
Cost per 1,000 queries$6.54$263$2,141$131
API price (input / output per 1M)$0.25 / $1.00$0.50 / $2.00$10 / $50$4 / $20
NDCG, time and cost rows are measured by SID on its own benchmark. Retrieval backend time is excluded.
How can developers deploy SID-2?
SID-2 is available through SID’s API and on Baseten. SDKs exist for Python (sid-sdk), TypeScript (@sid-ai/sid-sdk) and Go (github.com/sidhq/sid-go). A SageMaker helper covers AWS deployments. The reference implementation uses the OpenAI client format.
SID-2 costs $0.25 per million input tokens and $1.00 per million output tokens. Swarm costs $0.50 and $2.00. Cached input is 10% of the input price.
Key Takeaways
SID-2 is an RL-trained agentic search model that returns ranked document IDs.
SID-2 Medium matches Opus 5.5 retrieval at 0.585 NDCG in 3.11 s.
12-agent SID-2 Swarm leads SID’s benchmark at 0.720 NDCG.
Pricing starts at $0.25 per million input tokens, with no reindexing.
All results are self-reported; size and weights are undisclosed.
Check out the SID-2 technical report, the docs and the model page.