Signal mix
Developer mentions captured per vendor in this snapshot.
Surface-by-surface comparison
Click any surface to see how developers talk about all three vendors side-by-side.
Playground & API Console
The surface where developers iterate on prompts, test parameters, and generate code snippets before hitting the SDK.
Docs & Task-Oriented Guides
Depth and discoverability of documentation — cookbook recipes, framework integrations, error reference, versioning.
Enterprise Deployment & Data Residency
Self-hosted, VPC, multi-cloud (AWS Bedrock, Azure, Oracle), data residency guarantees, compliance certifications.
Pricing & Rate Limits
Per-token pricing clarity, free tier, rate limits, cost predictability for production workloads.
SDK Onboarding & First Call
How fast can a developer go from zero to first successful API call — install, auth, error handling, working example.
Model Selection Clarity
Developers asking 'which model should I use' — pricing vs capability tradeoffs, migration paths, deprecation signals.
How I built this
Same architecture I've used for two other PM applications — forked and pointed at Cohere.
Sources
- Curated21
Approach
- Six surfaces scoped from Cohere's Platform Experience & Developer Product JD
- Public signals from Reddit, Hacker News, GitHub Issues, Stack Overflow
- Keyword-cluster matching to surface where each vendor dominates
- Cohere stance scored relative to OpenAI + Anthropic on each surface
- Refreshable via authenticated POST to
/api/scrape