RXed AI News

AI to the bone.
Audited 2026-07-05 · RXed table v1.0

ChatGPT / OpenAI Platform

Visit openai.com
“Chat, work, create, code with AI” — the vendor’s own words

The deepest stack in AI and still the default — just budget for model churn and don't build on parts already marked for deprecation.

Best for: Teams that want one vendor covering the whole AI stack at frontier quality and accept vendor lock-in plus fast product churn.
Scope18/20
Quality8/10
Where the quality sits
8Reactive
8Retrieval & Memory
8Orchestration
6Validation
9Models
Frontier labChatAutomation & AgentsResearchFreemium
Vendor
OpenAI · openai.com
Origin
US — San Francisco
Pricing
Free · Go $8/mo · Plus $20/mo · Pro $100/mo · Business $25/mo · Enterprise Custom
Users (official only)
900M weekly active users, 50M paying subscribers (source, 2026-02-27)
FreeFreemay include ads
Go$8/momay include ads
Plus$20/mo
Pro$100/mofrom; $200/mo for 20x usage
Business$25/moper user, 2 seats min, $20/seat annually
EnterpriseCustomannual, contact sales

OpenAI Platform (API) access is billed separately pay-per-token and no ChatGPT subscription includes API credits.

checked 2026-07-25 · vendor pricing page

Element scores

Reactive
Retrieval & Memory
Orchestration
Validation
Models
Primitives
Pr8
Prompts
Em8
Embeddings
Cx9
Context
Tr7
Tracing
Lg9
LLM
Compositions
Fc9
Function calling
Vx7
Vector store
Rg8
RAG
Gr7
Guardrails
Mm9
Multimodal
Deployment
Ag8
Agents
Ft9
Fine-tuning
Fw7
Frameworks & harnesses
Ev5
Evaluations
Sm8
Small models
Emerging
Ma7
Multi-agent
Sy
Synthetic data
Pc7
Protocols
In
Interpretability
Th9
Thinking models
Tap or hover any element to see why it got that score.

Strengths

GPT-5.x is frontier-grade across chat, Codex, agents and Computer Use, and the platform covers nearly the full periodic table: hosted vector stores and file search, strong embeddings, the richest fine-tuning menu (SFT, vision, DPO, RFT), and an Agents SDK with built-in tracing, guardrails and MCP support.

Honest dings

The hosted Evals platform shuts down 30/11/2026 (the Ev score reflects that), Agent Builder ships with only ~8 working out-of-box integrations and locks you to OpenAI models, and 2026 brought ads to Free/Go tiers plus persistent complaints about cost-optimized routing and more refusals. Model churn is a real integration tax.

Prices and details change — this passport is re-verified at least quarterly.
Sources (7) — every claim traceable

Every audit lists the research it rests on — transparency and traceability are the product. Tools evolve: each audit is a snapshot of its audit date, and re-audits supersede older versions (kept below for reference).