Preference Model B Verified
RL environment company · ML
- Funding
- $16M raised · rounds and investors
- Tier
- B why
- Snapshot

What Preference Model does
Preference Model builds hardened RL environments with robust graders for frontier labs to train capable, better-aligned models, and open-sources its Karotte environment framework.
- Sells
- RL environments
- Leads with
- Several domains (a generalist)
- Based in
- San Francisco, USA
- Team
- 11-25
- Website
- preferencemodel.com ↗
What Preference Model sells
Read from Preference Model's own website, checked October 12, 2026: the line that puts it in each category.
RL environments
“That's why we're building RL environments for training capable, better-aligned superintelligences.”
preferencemodel.com ↗
Domains Preference Model covers
“Write CUDA kernels for the sliding-window attention layer of Inkling, Thinking Machines Lab's open-weights model, which no public kernel covers.”
preferencemodel.com ↗“Play the real Slay the Spire through game tools alone: pick a route through each act's map, win turn-based card fights”
preferencemodel.com ↗“Post-train Qwen3-4B, an open reasoning model, to solve competition math problems with far less reasoning, without losing accuracy.”
preferencemodel.com ↗
Preference Model competitors and alternatives
The companies that sell the most of the same things in the same domains, and what they share with Preference Model.
- Patronus AI Also sells RL environments · also in coding and games
- Prime Intellect Also sells RL environments · also in coding and math
- ReasonCore Also sells RL environments · also in math and coding
- Surge AI Also sells RL environments · also in coding and math
- Turing Also sells RL environments · also in coding and math
Tier 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 11-25
- raised $16M
- investors: a16z, Barry McCardel, Dylan Patel
- in curated RL directories (3 pts)
On X · mixed, moderate buzz, from 12 posts about Preference Model in the last month
Its Terminal-Bench audit drew praise, but Snorkel found ~55% of its fixes flawed; it also raised a $16M seed and open-sourced Karotte.
How we got to tier B
- Adoption: team 11-25; raised $16M; investors: a16z, Barry McCardel, Dylan Patel. Tier B.
- On X: mixed, moderate buzz. “Its Terminal-Bench audit drew praise, but Snorkel found ~55% of its fixes flawed; it also raised a $16M seed and open-sourced Karotte.” No move.
- In curated RL directories: 3 points, a floor of none.
- 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.
Preference Model funding
- Raised
- $16Msum of the rounds below
- Latest round
- Seed $16M2026-10 · led by a16z
Backed bya16zManifundMoE CapitalScale Angel GroupSignalFire
Rounds
- Seed · $16M · 2026-10 · led by a16z
we're backed by $16M in seed funding led by a16z
- Read: “we're backed by $16M in seed funding led by a16z”
- Found word for word on its own site: preferencemodel.com/blog/building-rl-environments-for-superintelligence
- The quote states the amount and who led it
- Also named by 1 other page read
Funds 5
- a16z led a round
we're backed by $16M in seed funding led by a16z, with participation from SignalFire, South Park Commons, Scale Angel Group
- Read: “we're backed by $16M in seed funding led by a16z, with participation from SignalFire, South Park Commons, Scale Angel Group”
- Found word for word on its own site: preferencemodel.com/blog/building-rl-environments-for-superintelligence
- Also named by 2 other pages read
- Manifund
South Park Commons, Scale Angel Group, Manifund, MoE Capital, Fei-Fei Li
- Read: “South Park Commons, Scale Angel Group, Manifund, MoE Capital, Fei-Fei Li”
- Found word for word on its own site: preferencemodel.com
- No other page or directory names it yet
- MoE Capital
South Park Commons, Scale Angel Group, Manifund, MoE Capital, Fei-Fei Li
- Read: “South Park Commons, Scale Angel Group, Manifund, MoE Capital, Fei-Fei Li”
- Found word for word on its own site: preferencemodel.com
- No other page or directory names it yet
- Scale Angel Group
we're backed by $16M in seed funding led by a16z, with participation from SignalFire, South Park Commons, Scale Angel Group
- Read: “we're backed by $16M in seed funding led by a16z, with participation from SignalFire, South Park Commons, Scale Angel Group”
- Found word for word on its own site: preferencemodel.com/blog/building-rl-environments-for-superintelligence
- No other page or directory names it yet
- SignalFire
we're backed by $16M in seed funding led by a16z, with participation from SignalFire, South Park Commons, Scale Angel Group
- Read: “we're backed by $16M in seed funding led by a16z, with participation from SignalFire, South Park Commons, Scale Angel Group”
- Found word for word on its own site: preferencemodel.com/blog/building-rl-environments-for-superintelligence
- No other page or directory names it yet
Accelerators 1
- South Park Commons
we're backed by $16M in seed funding led by a16z, with participation from SignalFire, South Park Commons, Scale Angel Group
- Read: “we're backed by $16M in seed funding led by a16z, with participation from SignalFire, South Park Commons, Scale Angel Group”
- Found word for word on its own site: preferencemodel.com/blog/building-rl-environments-for-superintelligence
- No other page or directory names it yet
Angels 8
- Barry McCardel
Julian Schrittwieser, Sammy Sidhu, Barry McCardel, Dylan Patel, and swyx.
- Read: “Julian Schrittwieser, Sammy Sidhu, Barry McCardel, Dylan Patel, and swyx.”
- Found word for word on its own site: preferencemodel.com
- No other page or directory names it yet
- Dylan Patel
Julian Schrittwieser, Sammy Sidhu, Barry McCardel, Dylan Patel, and swyx.
- Read: “Julian Schrittwieser, Sammy Sidhu, Barry McCardel, Dylan Patel, and swyx.”
- Found word for word on its own site: preferencemodel.com
- No other page or directory names it yet
- Fei-Fei Li
Manifund, MoE Capital, Fei-Fei Li, Ian Goodfellow, Jacob Jackson
- Read: “Manifund, MoE Capital, Fei-Fei Li, Ian Goodfellow, Jacob Jackson”
- Found word for word on its own site: preferencemodel.com
- No other page or directory names it yet
- Ian Goodfellow
Manifund, MoE Capital, Fei-Fei Li, Ian Goodfellow, Jacob Jackson
- Read: “Manifund, MoE Capital, Fei-Fei Li, Ian Goodfellow, Jacob Jackson”
- Found word for word on its own site: preferencemodel.com
- No other page or directory names it yet
- Jacob Jackson
Fei-Fei Li, Ian Goodfellow, Jacob Jackson, Julian Schrittwieser, Sammy Sidhu
- Read: “Fei-Fei Li, Ian Goodfellow, Jacob Jackson, Julian Schrittwieser, Sammy Sidhu”
- Found word for word on its own site: preferencemodel.com
- No other page or directory names it yet
- Julian Schrittwieser
Fei-Fei Li, Ian Goodfellow, Jacob Jackson, Julian Schrittwieser, Sammy Sidhu
- Read: “Fei-Fei Li, Ian Goodfellow, Jacob Jackson, Julian Schrittwieser, Sammy Sidhu”
- Found word for word on its own site: preferencemodel.com
- No other page or directory names it yet
- Sammy Sidhu
Julian Schrittwieser, Sammy Sidhu, Barry McCardel, Dylan Patel, and swyx.
- Read: “Julian Schrittwieser, Sammy Sidhu, Barry McCardel, Dylan Patel, and swyx.”
- Found word for word on its own site: preferencemodel.com
- No other page or directory names it yet
- swyx
Julian Schrittwieser, Sammy Sidhu, Barry McCardel, Dylan Patel, and swyx.
- Read: “Julian Schrittwieser, Sammy Sidhu, Barry McCardel, Dylan Patel, and swyx.”
- Found word for word on its own site: preferencemodel.com
- No other page or directory names it yet
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.
Where Preference Model sits
Its domain on the map, live. Hover a logo for its name; click to open it. Open it on the full map
Latest from Preference Model on X
@preferencemodel ↗Next: browse all environments. Entry from the RL Research daily scrape of public sources, 2026-11-05 snapshot.