Rocket - Think it. Type it. Launch it.
Mini toolsChroma Context 1
By Chroma
Context-1 is a purpose-trained retrieval and search agent rather than a general chat model. Chroma describes it as a 20B model that uses a self-editing context mechanism to discard irrelevant passages during search, helping it sustain performance over long trajectories inside a bounded context window. On Chroma’s reported evaluations, it matches or exceeds much larger models on generated and public benchmarks, and the company says it is released with open weights plus the full data generation pipeline, making it both a production-oriented search subagent and a reproducible research release.
New
Text
Gen 3
Overview
Context-1 is Chroma’s 20B agentic search model trained as a self-editing search agent. It is designed to decompose complex queries, prune irrelevant context, and deliver high retrieval quality at lower latency and cost than much larger frontier models.
About Chroma
Chroma is an open-source (Apache 2.0), AI-native data infrastructure company providing a serverless vector database supporting vector, full-text, regex, and metadata search. Built on object storage, it powers retrieval-augmented generation (RAG) and AI applications, with 24K+ GitHub stars and 11M+ monthly downloads.
Industry:
Data Infrastructure and Analytics
Company Size:
2-10
Location:
San Francisco, California, US
Website:
trychroma.com
Tools using Chroma Context 1
No tools found for this model yet.
Last updated: March 27, 2026
