Paper Instruments
Paper Instruments (formerly The LLM Data Company) is a San Francisco AI company from Y Combinator's 2025 batch.
About
Paper Instruments (formerly The LLM Data Company) is a San Francisco AI company from Y Combinator's 2025 batch. Starting with evals and post-training data tooling, it now builds open-source frontier models for professional knowledge work, including the Ultramarine model family, the Feather agent harness and the open-source Paper Office document libraries, targeting finance, legal, consulting and operations professionals.
Capital graph
Founders / team· 3
Comparable companies· by market
This company
BusinessLatest roundTotal raisedDate- Paper Instruments
Post-trains and open-sources LLMs for knowledge work (Ultramarine), with the Feather agent tool…
—
United States
BusinessLatest roundTotal raisedDate- Celeris
Ultra-low-latency general-purpose diffusion-based large language model (Celeris-1), offered via an OpenAI-compatible API
— - OpenAI
GPT series of large language models and consumer and enterprise AI products such as ChatGPT
Undisclosed · $180B · 2026-03 · Valuation $852B
$180BValuation $852B - Inception
Diffusion large language models (DLMs) for text and code
Seed · $50M · 2025-11
$50M
China
BusinessLatest roundTotal raisedDate- DeepSeek
Open-source large language model development and API/agent services
Series A · $7.2B · 2026-06 · Valuation $56B
$7.2BValuation $56B - Mindverse
Post-training and continual learning for agent models…
Series A · $50M · 2026-06
$50M - Moonshot AI
Develops the Kimi series of large models and open-sources their weights; offers the Kimi intelligent assistant, Kimi Code and an open platform API
Pre-IPO · $8.0B · 2026-10 · Valuation $50B pre-money
$8.0BValuation $50B pre-money
Europe
BusinessLatest roundTotal raisedDate- Mistral AI
Develops open-weight large language models and provides supporting compute infrastructure…
Series D+ · 2026-09 · Valuation $24B
—Valuation $24B - Aleph Alpha
Develops AI large language models based on European data
Series B · $500M · 2023-11
$500M - H
Computer-use AI agent platform for large European enterprises.
Seed · $220M · 2024-05
$220M
Other / unspecified
BusinessLatest roundTotal raisedDate- UkisAI
Trains and open-sources efficient reasoning LLMs (Swift series), and provides custom AI solutions and AI training for enterprises
— - Sarvam AI
Builds large language models that support Indian languages
Series B · $275M · 2026-06 · Valuation $1.5B
$275MValuation $1.5B - Cohere
Platform company that builds general-purpose and vertical large models for enterprises
Series D+ · $1.8B · 2026-10 · Valuation $26B
$1.8BValuation $26B
Confirmed rounds only. Matched on subsector, sector, stage and funding scale — each verifiable on the company profiles.
Investors that may fit
Andreessen Horowitz
5 Foundation Models deals in the last 24 months · 4 in Silicon Valley
Request intro →Salesforce Ventures
4 Foundation Models deals in the last 24 months · 2 in Silicon Valley
Request intro →Ranked by each firm’s deals in this sector over the last 24 months, weighted by nearby stage and shared geography. Indicative only — no rating judgment.