Collinear B Verified

RL environment company · Security

Domains
Security Enterprise Workflows Coding Computer Use Long-Horizon
On X
@collinearAI ↗
Tier
B why
Snapshot

Open environment ↗collinear.ai

What Collinear does

Collinear sells Simulation Lab RL environments with simulated users, tools and verifiers, plus off-the-shelf tasks, benchmarks and agent training data for frontier labs.

Leads with
Several domains (a generalist)
Based in
Mountain View, USA
Team
11-25

What Collinear sells

Read from Collinear's own website, checked October 11, 2026: the line that puts it in each category.

RL environments

  • RL environments

    “Every simulation lab is a self-contained world where your agent operates, complete with the users, tools, data, and tasks it will face in production.”
    collinear.ai/ ↗

  • Evals

    “Diagnose failure, don’t just rank models. Including CWE-Bench, MCP-Atlas, and OSWorld-Verified”
    collinear.ai/ ↗

Data

Domains Collinear covers

Benchmarks Collinear built or runs

Collinear competitors and alternatives

The companies that sell the most of the same things in the same domains, and what they share with Collinear.

Every RL environment company on the landscape map →

Tier B

B

Growing: 11 to 50 people, or $3M+ raised.

How much stands behind a tier: usage, who backs it, a confirmed line from its own site or paper, and a clear read on X. High with three of the four, medium with two, low otherwise.

  • team 11-50
  • raised $10M
  • investors: Engineering Capital, 112 Capital, B Capital
  • in curated RL directories (5 pts)

On X · well regarded, low buzz, from 14 posts about Collinear in the last month

Mostly founder and staff posts claim its benchmarks are adopted by frontier labs; it is an Artificial Analysis cyber index partner, with little independent discussion.

@nazneenrajani ↗ · @nazneenrajani ↗ · @nazneenrajani ↗

How we got to tier B
  1. Adoption: team 11-50; raised $10M; investors: Engineering Capital, 112 Capital, B Capital. Tier B.
  2. On X: well regarded, low buzz. “Mostly founder and staff posts claim its benchmarks are adopted by frontier labs; it is an Artificial Analysis cyber index partner, with little independent discussion.” No move.
  3. In curated RL directories: 5 points, a floor of C.
  4. Confirmed: a line from its own site or paper places it.

A clear view on X moves a tier one step up or down; most entries don't move. How tiers work.

Collinear funding

Backed by112 CapitalB CapitalEngineering CapitalHitachi VenturesStorm Ventures

Funds 4

Corporate investors 1

Every figure is quoted from its own site, its accelerator or a funding announcement; a round's amount and lead appear only when its quote states them. Startup databases are never used as evidence. Checked 2026-10-11. How funding is checked.

Training methods

Methods Collinear uses, offers or builds for, with public evidence: its own website says so, or two independent sites do.

  • DPO Direct Preference Optimization On its own website

    “4. Preference Optimization: Iteratively improve the judge model by DPO, resulting in an improved judge model at each iteration.”

    blog.collinear.ai
  • SFT Supervised Fine-Tuning On its own website

    “…ods, by relying only on examples from the training data, our model significantly improves both its rationale quality and its scoring accuracy, outperforming similar and even bigger sized models as well as different post-SFT methods, on diverse evaluation benchmarks. Additionally, we also show the quality improvement in rationales through our proposed recipe through human evaluation.”

    blog.collinear.ai

    “…12B-23B to match the performance of models in the range of 32B-70B. This blog presents results on using Collinear math correctness curators used by ServiceNow for CPT of Mistral-NeMO-12B. We also show similar gains with SFT on a Phi-3.5B-instruct model using their curated data. The curated datasets and the post-trained models are now open-sourced and available on Hugging Face.”

    blog.collinear.ai

Where Collinear sits

Its domain on the map, live. Hover a logo for its name; click to open it. Open it on the full map

The galaxy could not load. See every vendor on the board.

Latest from Collinear on X

@collinearAI ↗

More Security RL environments

Security AI benchmarks

Next: browse all environments. Entry from the RL Research daily scrape of public sources, 2026-11-05 snapshot.