Drop-in OpenAI-compatible API with persistent lattice memory. Crystallize documents once into permanent Lattice Nodes and recall Grounded Facts through a compact scoped viewport. Official-scale results: 96.28% exact (466/484) on RULER MRCR v2 and 54.87% (276/503) on LongBench v2.
Paired evaluation used Google Gemini 3.5 Flash-Lite (gemini-3.5-flash-lite) as the upstream BYOK test model. ICX works with any supported LLM — Gemini was the model we ran for this comparison. Same tasks, same weights — ICX separates permanent lattice storage from the generative viewport so the model receives only the facts it needs. Token cuts below are versus uncached full-context Gemini prefill. They are not a dollar comparison to a prompt-cached Cursor, Claude Code, or Antigravity harness.
gemini-3.5-flash-lite · BYOK via Google AI API
LongBench v2 paired control · n=10 · identical item IDs
RULER MRCR · 484 rows · multi-round ordinal disambiguation
ICX run 073110 · Gemini full-context 20260819_104413 · same 503 IDs
852k context · deep AST & cross-file · item 66fa208b
| Benchmark & workload | Gemini 3.5 Flash-Lite alone | Gemini 3.5 Flash-Lite + ICX | Advantage |
|---|---|---|---|
| LongBench v2 paired n=10 matched item IDs |
6/10 · 2,253,544 tokens | 6/10 · 162,923 tokens | 92.77% vs uncached prefill · n=10 only |
| RULER MRCR 484 rows · eval 9fc0e405 |
No 484-row Gemini dump in archive | 96.28% (466/484) | 18 ordinal sibling misses |
| LongBench v2 official 503 items |
38.57% (194/503 raw) · 188 API failures | 54.87% (276/503) | +16.30 pp raw · 0 ICX errors |
| SWE-bench Verified 500 · predicted pass |
No official test-resolved column | 96.40% (482/500) | Not official test-resolved |
| Code repository 852k context |
6,227 ms · 852,861 tokens | 1,375 ms · 31,783 tokens | 4.53× faster · 88.81% token cut |
Prompt caching discounts re-reading a stable prefix. It does not take those tokens off the model’s working desk. ICX keeps the archive in lattice memory and streams a scoped viewport.
Everything your engineering, security, and product teams need to deploy infinite context memory with zero compliance or performance friction.
Bring Your Own Google Gemini or OpenAI API keys. Keys are forwarded ephemerally via request headers (X-LLM-API-Key) with zero server-side key retention or training on customer data.
100% compatible drop-in replacement for standard OpenAI SDKs in Python, Node.js, Go, and Rust. Simply update baseURL and pass X-Space-ID to begin grounding requests.
Native 1-click Model Context Protocol (MCP) server integration for Google Antigravity, Claude Desktop, Cursor, and custom autonomous agent swarms.
Explore Hosted MCP Docs →Multi-tenant namespace isolation (legal_vault, eng_repo, clinical_ehr) ensures zero cross-tenant factual leakage and instantaneous context switching.
Discrete simplicial graph diffusion executes on standard CPU memory in <0.5ms, eliminating expensive GPU cluster reservation fees and VRAM memory starvation.
Factual assertions form closed simplicial chains (∂c = 0). Hallucinated or out-of-ontology statements trip the boundary refusal operator instantly.
Drop Calera ICX into your existing OpenAI, LangChain, LlamaIndex, or REST workflows in less than 60 seconds.
Want a live comparison before copying code? Open the ICX console preview or run a FinanceSec query — both use the same Calera account, billed per product.
import requests from openai import OpenAI API_KEY = "icx_live_your_api_key" BASE_URL = "https://icx.api.caleralabs.com/v1" SPACE_ID = "legal_vault" # 1. Ingest document into isolated memory space via REST API with open("contracts/vendor_agreement.pdf", "rb") as f: requests.post( f"{BASE_URL}/ingest/file", headers={"Authorization": f"Bearer {API_KEY}", "X-Space-ID": SPACE_ID}, files={"file": f} ) # 2. Query associative memory via standard OpenAI client drop-in client = OpenAI( api_key=API_KEY, base_url=BASE_URL, default_headers={"X-Space-ID": SPACE_ID} ) response = client.chat.completions.create( model="calera-icx-v1", messages=[ {"role": "user", "content": "Recall exact contractual indemnity cap across all MSAs."} ] ) print(response.choices[0].message.content)
A formula, not an invoice. It applies a published list price and a published cache-read multiplier. ICX has not measured a live agent-loop bill.
Assumptions (all modeled): input $2.50 / MTok; Anthropic cache-read 0.10× if the entire archive were a stable prefix; ICX viewport 1,000 tokens/turn + the matching subscription. A prefix above ~1M tokens does not fit in one window, so the cache-read column is algebra, not a feasible agent turn.
No savings percentage is shown. This sketch is not a measured invoice and is not a win against a warm cache.
Read Infinite Memory Is Not Infinite Attention (CALERA-PAPER-ICX-2026-08-23). Named-run scores: RULER 466/484, LongBench 276/503 vs Gemini 194/503 raw, SWE-bench Verified 482/500 predicted pass. Row dumps are not published on that page.
Calera asked Google Antigravity, with the ICX MCP attached to this repository, to contrast truncated grep with lattice lookup. The result is an operator transcript — not a measurement. Empirical head-to-head numbers stay on the paper above.
Read the sessionSelect a developer preview tier. All subscriptions include single-account management and multi-domain billing on Calera Labs.
For individual developers testing persistent memory workflows.
For solo developers who want persistent memory without a team invoice.
For engineering teams building memory-augmented AI applications.
For production AI workloads and heavy context applications.
Self-hosted + unlimited + white-glove. Contact sales.
Provision isolated, high-throughput Cloud Run worker instances with guaranteed 1,500 req/min burst rates, dedicated CPU/memory specs, and private memory space isolation for enterprise security requirements.