No topics yet. Start the conversation.

Summary

User supplied summary for the plot

Count of publicly announced plan-tightening actions per vendor, cumulative.

Description

The below description is supplied in free-text by the user

The canary thesis

Agentic coding sessions burn 10–100× more tokens per user than chat did, so every vendor selling a flat-rate seat is quietly losing money on heavy users. The thesis: if every agentic vendor tightens within a 90-day window, flat-rate is dead. This chart tracks the canary events.

Signal strength: 34 tightening events across 7 vendors (Amazon, Anthropic, Cursor, GitHub Copilot, JetBrains, OpenAI, Windsurf) since the first move in May 2025. The March-April 2026 cluster (Windsurf overage + OpenClaw block) is a second clear tightening window.

Event log

Full timeline of every pricing, metering, and harness-restriction action we track. Tightens = Y means the action transfers cost or risk from vendor to user. Pre-2025 rows are baseline context and don't count toward the chart's cumulative lines.

Date Vendor Tightens Headline
2022-06-21 GitHub Copilot Y GitHub Copilot becomes paid product at $10/mo individual
2023-02-14 GitHub Copilot Y Copilot for Business launches at $19/mo
2024-02-27 GitHub Copilot Y Copilot Enterprise launches at $39/mo
2024-04-04 OpenAI Y OpenAI launches Batch API with 50% discount
2024-04-30 Amazon Y Amazon Q Developer launches (CodeWhisperer rebrand) with Free + $19 Pro tiers
2024-08-14 Anthropic Y Anthropic launches prompt caching (0.1× read, 1.25× write)
2024-10-09 Anthropic Y Anthropic Batch API launches with 50% off input + output
2024-11-01 Cursor Y Cursor Pro at $20/mo with 500 fast + unlimited slow requests
2024-11-13 Windsurf Y Codeium launches Windsurf Editor (first agentic IDE)
2024-12-05 OpenAI Y ChatGPT Pro $200/mo launches — first major $200 consumer AI tier
2025-01-01 OpenAI Y OpenAI Batch API input/output prices drop further for GPT-4o family (community-observed)
2025-04-01 Anthropic Y Anthropic adds 1-hour cache write tier at 2× base
2025-04-09 Anthropic Y Claude Max $200/mo (20× Pro) launches; $100/mo (5×) added later
2025-04-22 Windsurf Y Codeium company rebrands to Windsurf
2025-05-07 GitHub Copilot Y Copilot announces enforcement of premium request limits (moved to June 4)
2025-05-13 Windsurf Y Reports of OpenAI $3B deal to acquire Windsurf surface
2025-06-16 Cursor Y Cursor switches Pro from request-based to usage-based (token-cost) billing
2025-06-17 Cursor Y Cursor Ultra $200/mo (20× Pro) launches
2025-06-18 GitHub Copilot Y Premium requests billing goes live across paid Copilot plans
2025-07-04 Cursor Y Cursor apologises and refunds unexpected charges from June pricing rollout
2025-07-28 Anthropic Y Anthropic announces weekly rate limits for Claude Pro/Max, effective August 28
2025-08-01 GitHub Copilot Y Premium requests billing extends to GHE.com
2025-08-15 Cursor Y Teams moves from fixed per-request to variable request costs; Auto usage adjusted
2025-08-25 JetBrains Y JetBrains reduces AI Credits quotas for Pro and Ultimate under new transparent quota model
2025-08-28 Anthropic Y Claude weekly rate limits take effect on Pro/Max plans
2025-12-02 GitHub Copilot Y Copilot removes $0 premium-request budget safety for enterprise/team accounts
2026-02-05 Anthropic Y Prompt cache isolation moves from org-level to workspace-level on Claude API
2026-02-13 OpenAI Y OpenAI details real-time access system for Sora and Codex with usage tracking and credits.
2026-03-19 Windsurf Y Windsurf raises Pro $15 to $20 and moves overage to API pricing
2026-04-02 OpenAI Y OpenAI introduces pay-as-you-go pricing for Codex in Business and Enterprise plans
2026-04-04 Anthropic Y Anthropic blocks OpenClaw from Claude Pro/Max subscriptions
2026-04-21 Anthropic Y Anthropic briefly removes Claude Code from $20 Pro plan (later reversed)
2026-04-27 GitHub Copilot Y GitHub Copilot code review to consume GitHub Actions minutes
2026-04-27 GitHub Copilot Y Copilot Student Plan: GPT-5.3-Codex removed from direct model selection

Source: vendor blogs, changelogs, and docs. Stored in Turso ai-inference-unit.events; refreshed nightly by the ai-inference-unit-refresh Novem job.