Locked Population Definition and Verification Gates
The AI Dev-Tools Census is a global product-category reference built for software engineers, engineering managers, and platform architects evaluating developer tools. Every profile in this census is strictly non-promotional: there are no rankings, no sponsored placements, no comparison scores, and no unverified vendor claims. Every figure traces to a live primary company page verified on November 20, 2026.
- Population definition: Companies whose primary product is (a) an AI coding agent or coding assistant, (b) Model Context Protocol (MCP) servers, clients, registries, or infrastructure tooling, or (c) enterprise AI developer tooling.
- Gate 1 — Independent: Must operate as an independent company rather than an acquired business unit or subsidiary product line. Live verification in November 2026 pruned major acquired products including Cursor (acquired by SpaceX, August 2026), Windsurf (acquired by Cognition, July 2025), Augment Code (core products acquired by Harness, October 2026), Tabnine (acquired by Tricentis, July 2026), Langfuse (ClickHouse, January 2026), Arize AI (Dynatrace, October 2026), Weights & Biases (CoreWeave, May 2025), Helicone (Mintlify, March 2026), Portkey (Palo Alto Networks, 2026), Smithery (Arcade, August 2026), Tavily (Nebius, February 2026), and Neon (Databricks, May 2025).
- Gate 2 — Operating: Product must be live, obtainable, and free of active shutdown or dead-endpoint wind-downs.
- Gate 3 — In-population: Primary product must match the locked definition; general frontier AI platforms with a coding feature (such as OpenAI or Anthropic) and general SaaS/cloud/DevOps suites that merely expose an MCP server are excluded from publication.
85-Company Final Reconciliation Summary
| Disposition | Count | Summary |
|---|---|---|
| PUBLISH (Chunk 1 Built) | 30 | First 30 companies in CSV rank order (#7 to #77) passing Independent, Operating, and In-population gates. |
| PUBLISH (Chunk 2 Built) | 4 | 3 queued Chunk 1 clearances (#78 Spiderbrain, #82 usefulapi.io, #85 XenMCP) + 1 HOLD resolved to PUBLISH (#83 UGT). |
| PRUNE | 43 | Failed Independence (16 acquired/subsidiary), Operating (1 dead URL), Non-company repo/actor listing (3), duplicate (1), or primary-product In-population gate (22). |
| HOLD | 1 | Operating preview whose primary /terms & /privacy explicitly state production operator/jurisdiction/paid terms are not finalized (#74 PlaywrightGen). |
| WATCH | 7 | Unsettled primary-product boundary, rebrand, or reverse-acquihire status (#10 Vercel, #15 Poolside, #17 JetBrains, #38 Upstash, #52 Convex, #59 n8n, #67 clauxel.com). |
| Total Gated Population | 85 | 34 Published + 43 Pruned + 1 Hold + 7 Watch = 85 (100% reconciled as of November 20, 2026). |
Published Profiles (34 Companies in CSV Rank Order)
All 34 independent, operating, in-population companies that cleared verification across Chunk 1 (30 profiles, ranks #7 through #77) and Chunk 2 (4 profiles: #78 Spiderbrain, #82 usefulapi.io, #83 UGT, and #85 XenMCP) are listed below in exact CSV rank order.
| # | CSV Rank | Company Profile | Cluster | Published Pricing Summary | Artifacts & Source |
|---|---|---|---|---|---|
| 1 | #7 | Cognition (Devin) | AI Coding Agents & Editors | Free ($0), Pro ($20/mo), Max ($200/mo), Teams ($80/mo, up to 100 users), Enterprise (custom) | .md Twin · Official Site |
| 2 | #8 | Sourcegraph | AI Coding Agents & Editors | Enterprise plan starting at $16,000/year minimum annual contract (scales with team size and AI credit buckets) | .md Twin · Official Site |
| 3 | #9 | Replit | AI Coding Agents & Editors | Core ($20/mo monthly or $18/mo billed annually), Pro (from $100/mo or $90/mo annual with higher credit tiers up to $2,000/mo), Enterprise (custom) | .md Twin · Official Site |
| 4 | #11 | StackBlitz (Bolt.new) | AI Coding Agents & Editors | Bolt.new Free ($0/mo, 1M tokens/mo) and Pro (starting at $25/mo, 10M tokens/mo); StackBlitz Personal free + Teams/Enterprise | .md Twin · Official Site |
| 5 | #12 | Lovable | AI Coding Agents & Editors | Free ($0), Pro ($25/mo, 100 monthly credits), Business ($50/mo, 100 monthly credits + SSO), Enterprise (custom) | .md Twin · Official Site |
| 6 | #14 | Factory AI | AI Coding Agents & Editors | Free ($0 BYO inference), Pro ($20/mo), Plus ($100/mo), Max ($200/mo), Teams & Enterprise (custom) | .md Twin · Official Site |
| 7 | #16 | Zed Industries | AI Coding Agents & Editors | Personal ($0 forever, unlimited BYOK/external agents), Pro ($10/mo, unlimited edit predictions + $5 tokens), Business ($30/seat/mo) | .md Twin · Official Site |
| 8 | #19 | Qodo | AI Coding Agents & Editors | Free 14-day trial; Pro Team from $30/mo ($0.015/credit pooled, up to 30 users); Enterprise custom (30+ users); free for open source | .md Twin · Official Site |
| 9 | #20 | Cline | AI Coding Agents & Editors | Open Source Free ($0 software; BYO API keys or at-cost usage inference); Enterprise (custom pricing) | .md Twin · Official Site |
| 10 | #22 | Warp | AI Coding Agents & Editors | Free ($0/mo with BYO API keys & pay-as-you-go), paid agent tiers ($20/mo and $200/mo monthly; $18/$180 annual), Business ($50/user/mo), Enterprise (custom) | .md Twin · Official Site |
| 11 | #29 | LangChain | Agent Engineering, Observability & Infrastructure | LangSmith Developer ($0/seat, 5k base traces/mo), Plus ($39/seat/mo, 10k traces + Agent Server), Enterprise (custom); LSU usage billing ($1.00/LSU) | .md Twin · Official Site |
| 12 | #32 | Braintrust | Agent Engineering, Observability & Infrastructure | Starter ($0/mo, unlimited users, 1 GB processed data, 10k scores), Pro ($249/mo, 5 GB, 50k scores), Enterprise (custom) | .md Twin · Official Site |
| 13 | #36 | Composio | Agent Engineering, Observability & Infrastructure | Free (up to 3 members), Standard ($18/seat/mo including $18/seat credits + SSO & Firewall), Enterprise (custom with SCIM & BYOC) | .md Twin · Official Site |
| 14 | #39 | Mintlify | Agent Engineering, Observability & Infrastructure | Free ($0/mo, 5 editors + MCP server), Pro ($450/mo annual or $540/mo monthly, unlimited editors), Enterprise (custom) | .md Twin · Official Site |
| 15 | #40 | E2B | Agent Engineering, Observability & Infrastructure | Hobby (Free + $100 one-time credits + per-second usage), Pro ($150/mo + per-second usage), Enterprise (custom + BYOC) | .md Twin · Official Site |
| 16 | #41 | Modal | Agent Engineering, Observability & Infrastructure | Starter ($0 + per-second compute, $30/mo free credits); pay-per-second GPU (T4 $0.000164/s to B300 $0.001972/s) & CPU ($0.0000131/core/s) | .md Twin · Official Site |
| 17 | #42 | Daytona | Agent Engineering, Observability & Infrastructure | Pay-as-you-go per second ($200 free compute included); vCPU $0.0504/h, RAM $0.0162/GiB/h, GPUs from $0.57/h preemptible | .md Twin · Official Site |
| 18 | #43 | Browserbase | Agent Engineering, Observability & Infrastructure | Free ($0, 3 concurrent browsers), Developer ($20/mo, 25 concurrent, 100 hrs), Startup ($99/mo, 100 concurrent, 500 hrs), Enterprise (custom) | .md Twin · Official Site |
| 19 | #44 | Firecrawl | Agent Engineering, Observability & Infrastructure | Free ($0, 1k credits/mo), Hobby ($16/mo annual, 5k), Standard ($83/mo annual, 100k), Growth ($333/mo annual, 500k), Scale/Enterprise | .md Twin · Official Site |
| 20 | #45 | Exa | Agent Engineering, Observability & Infrastructure | Starter Free ($10/mo credits + $10 onboarding bonus), Developer PAYGO (Search from $4/1k requests, SOC 2 Type II), Enterprise (custom) | .md Twin · Official Site |
| 21 | #53 | Pinecone | Vector Retrieval & Multi-Agent Orchestration | Starter (Free), Builder ($20/mo flat), Standard ($50/mo min PAYGO + 3-week $300 trial), Enterprise (custom + BYOC) | .md Twin · Official Site |
| 22 | #54 | Weaviate | Vector Retrieval & Multi-Agent Orchestration | Free Forever ($0/mo, 100k objects, 1 GB RAM), Flex (from $45/mo PAYGO, 99.5% uptime), Premium (custom shared/dedicated) & self-managed OSS | .md Twin · Official Site |
| 23 | #55 | Qdrant | Vector Retrieval & Multi-Agent Orchestration | Free Forever Cloud (1 GB RAM / 4 GB disk), Standard usage-based (99.5% SLA), Premium (minimum spend, 99.9% SLA, SSO/VPC), Hybrid Cloud & OSS | .md Twin · Official Site |
| 24 | #56 | LlamaIndex | Vector Retrieval & Multi-Agent Orchestration | Free (10,000 credits/mo), Pay-as-you-go ($1.50/1k credits after free tier), Pro (400k credits/mo, $1.25/1k overage), Enterprise (custom) | .md Twin · Official Site |
| 25 | #57 | CrewAI | Vector Retrieval & Multi-Agent Orchestration | Basic (Free, 50 workflow executions/mo + visual Studio) and Enterprise (Custom, VPC/self-hosted, SSO/RBAC/PII redaction) | .md Twin · Official Site |
| 26 | #58 | Pydantic | Vector Retrieval & Multi-Agent Orchestration | Logfire Personal Free forever (10M telemetry records/mo, $20 value), then $2 per million records on Team/Growth; Enterprise custom | .md Twin · Official Site |
| 27 | #65 | AgentDiet | MCP Infrastructure, Auditors & Registries | Free Sandbox ($0), Pay-Per-Call x402/MPP ($0.002/call), Pro Developer ($29/mo, 20k calls), Enterprise SLA ($199/mo, 150k calls) | .md Twin · Official Site |
| 28 | #70 | Kaiva | MCP Infrastructure, Auditors & Registries | Free ($0 forever, 10 MCP servers, 200k tool calls/mo), Pro ($19/mo, unlimited servers, 1M calls then $0.04/1k), Enterprise (custom) | .md Twin · Official Site |
| 29 | #75 | PulseMCP | MCP Infrastructure, Auditors & Registries | Free public directory and newsletter; commercial partnerships/sponsorships via Work With Us inquiry (no published self-serve SaaS pricing) | .md Twin · Official Site |
| 30 | #77 | Robinson AI Systems | MCP Infrastructure, Auditors & Registries | Free public beta across live MCP tools (no paid checkout published as of November 20, 2026) | .md Twin · Official Site |
| 31 | #78 | Spiderbrain | Agent Engineering, Observability & Infrastructure | Beta API access on spiderbrain.ai/api/; dedicated single-tenant server deployments via sales on spiderbrain.ai/business/ | .md Twin · Official Site |
| 32 | #82 | usefulapi.io | MCP Infrastructure, Servers, Gateways & Registries | Self-serve hosted MCP connector endpoints on usefulapi.io; no separate public paid tier table published | .md Twin · Official Site |
| 33 | #83 | UGT (Use the Geek Thing) | MCP Infrastructure, Servers, Gateways & Registries | Free and open-source under the Apache License 2.0 (per usegeekthing.com/terms) | .md Twin · Official Site |
| 34 | #85 | XenMCP | MCP Infrastructure, Servers, Gateways & Registries | Free during beta (fair call limits, no credit card); planned post-beta Free and Pro tiers not yet finalized | .md Twin · Official Site |
Census Closeout: Watch (7) and Hold (1) Cases
Consistent with the census stop-and-report policy, cases where current primary sources do not settle independence, operator identity, or the primary-product boundary are closed transparently rather than forced into print:
- WATCH (7 cases): #10 Vercel (frontend/Next.js cloud hosting vs. Agentic Infrastructure revenue split undisclosed), #15 Poolside (operates poolside.ai and Laguna 2.1 models while August 2026 Nvidia deal transferred 100+ engineers and non-exclusive Model Factory license), #17 JetBrains (rolling out JetBrains Air / retiring standalone Junie plugin Dec 2026 while core suite remains multi-language IDEs), #38 Upstash (serverless Redis/Vector platform vs. Context7 MCP revenue split undisclosed), #52 Convex (reactive TypeScript BaaS vs. AI-agent backend primary-product split undisclosed), #59 n8n (business workflow automation vs. AI developer tooling boundary undisclosed), and #67 clauxel.com (redirects to coachix.dev personal AI assistant membership; /terms and /privacy return 404).
- HOLD (1 case): #74 PlaywrightGen (live /terms and /privacy explicitly state it is a protected preview whose production operator, jurisdiction, and paid terms are not yet finalized and checkout remains disabled). Note: #83 UGT cleared its Chunk 1 HOLD after primary-source checks on usegeekthing.com/terms, /privacy, /recipes, and GitHub verified operator Cameron Low / auscaster, Apache 2.0 free pricing, and local Linux engine architecture.
- Final Census Counts (November 20, 2026): 85 gated companies = 34 Published + 43 Pruned + 1 Hold + 7 Watch.
Get The Signal
What we actually ran, what it measured, and what it means for a restoration shop — every week, including the flops.

Leave a Reply