Claradb
Sign in
How it works

A quantified answer to 'is this company actually gaining traction?'

The Momentum Score is a composite signal built from 24+ underlying data sources. It combines code velocity, citation growth, funding pace, hiring patterns, model adoption, partnership activity, and benchmark performance into a single inspectable number.

Not a vanity metric. Every score is decomposable — click into any company and see which signals drove the number up or down, then decide whether you agree.

Code velocity
Research output
Funding signals
Market adoption
Team growth
Benchmark performance
Momentum Score
Anthropic
94
Code velocity
GitHub stars
287k (+12% 30d)
Commit freq
4.2 commits/day avg
Contributors
342 active (90d)
HF downloads
18.4M weekly
Research output
Papers (90d)
14 published
Citations
12,847 total
ArXiv rank
Top 3 in NLP
Patents filed
23 (trailing 12mo)
Funding & growth
Last round
Series D — $2B
Total raised
$7.3B
Headcount
1,200 (+34% YoY)
Key hires
VP Eng (ex-Google)
Trajectory

See momentum before the consensus narrative catches up.

Momentum scores update on a continuous cadence. Track how companies accelerate, plateau, or decline over weeks and months — not just at funding announcement time.

Identify inflection points before they become obvious: a spike in GitHub activity, a cluster of new papers, or a sudden hiring burst.

W12
Mixtral 8x22B released
+6.2
W08
Series B closed at $415M
+4.8
W03
3 papers accepted at NeurIPS
+2.1
Mistral AI
Past 16 weeks
87
+18.4
9586786960W01W02W03W04W05W06W07W08W09W10W11W12W13W14W15W16
The difference

Not another vanity metric.

Traditional scoring
  • Funding amount as a proxy for quality
  • Static snapshot updated quarterly
  • Opaque methodology behind a paywall
  • Single dimension (money raised or revenue)
  • Lagging indicator — you find out after the market moves
Momentum Score
  • Composite of code, research, adoption, and funding signals
  • Live scoring on a 24-hour update cadence
  • Inspectable methodology — see exactly how scores are built
  • Six scoring dimensions weighted by signal quality
  • Leading signals — GitHub, ArXiv, and hiring move before funding rounds
Data sources

24+ sources feed every score.

The Momentum Score is only as good as its inputs. Every source is documented, every weight is inspectable, and the full methodology is publicly available.

GitHub

Daily

Repository activity, star velocity, contributor growth, commit frequency, and release cadence across tracked organizations.

2,400+ org profiles

ArXiv

Daily

New paper submissions, citation tracking, author affiliation mapping, and cross-reference linking to company entities.

NLP, CV, ML categories

Hugging Face

Daily

Model downloads, dataset usage, space deployments, and community engagement metrics for tracked entities.

All public models & datasets

Patent databases

Weekly

New filings, grant status changes, citation networks, and technology classification mapping across USPTO and EPO.

AI-classified patents

Funding records

Real-time

Round announcements, valuation data, investor mapping, and follow-on patterns from public and syndicated sources.

Global AI & deep tech

Hiring signals

Weekly

Job postings, headcount changes, key hire announcements, and team composition shifts across tracked companies.

Engineering & research roles
See it live

Explore the live momentum rankings.

Browse the top AI companies by momentum score. Every company profile decomposes its score so you can see exactly what is driving the number.