What changed in AI tools: week 40, 2026
28 Sep – 4 Oct 2026 · 27 updates from 23 tools · 1 verified price or limit change. Pricing and limits first.
Pricing 5
- Kilo now runs on your ChatGPT plan across every surface
Sign in with ChatGPT and use your plan for inference in VS Code, JetBrains, the CLI, Cloud Agents, Code Reviews and the mobile app.
Source: blog.kilo.aiCurrent Kilo Code plans & limits - Devin Desktop (ex-Windsurf) can bill GPT requests to your ChatGPT plan
Sign in with ChatGPT to bill GPT model usage to an eligible Plus or Pro plan.
Source: docs.devin.aiCurrent Windsurf plans & limits - Warp's agent now runs on your ChatGPT plan
Sign in with ChatGPT to use your subscription's included Work and Codex usage in the Warp Terminal and Agent CLI.
Source: warp.devCurrent Warp plans & limits - Manus Flex lets you bring your own API key
Your provider bills inference directly, while other Manus services still use credits; OpenRouter, Fireworks and Modal are the first partners.
Source: manus.imCurrent Manus plans & limits - Character.ai adds (c.ai) lite at $4.99/month, plus two new free chat styles
Lite ($4.99/month or $2.49/week, varies by region) adds daily premium-style messages, no banner ads and less slow mode; PipSqueak 3 and ShortSqueak are free for everyone.
Source: blog.character.aiCurrent Character.AI plans & limits
Launch 8
- Qodo 3.0 targets quality and governance for agent-written code
Qodo pitches 3.0 as the quality and governance layer for an agentic software factory.
Source: qodo.ai - JetBrains Air comes to its IDEs in early access, with free Junie Lite runs
The plugin connects agents and subscriptions you already use, ships with no agents installed, and is in 2026.3 EAP builds.
Source: blog.jetbrains.com - Kiro workflows run multi-agent tasks from start to finish
Reusable multi-step agent plans arrived in the IDE, the CLI and (opt-in) Kiro Web cloud sessions.
Source: kiro.dev - Suno Speech beta generates spoken word and music as one track
The beta is now open to everyone after a month of limited testing.
Source: suno.com - HeyGen Video API: text, image and reference-to-video generation
The heygen-video-1 endpoints generate 5–15 second clips at 480p or 768p from prompts of up to 5,000 characters.
Source: docs.heygen.com - Fireflies Talk adds free, unlimited voice dictation to the desktop app
Hold a key and speak, and clean text appears wherever your cursor is.
Source: fireflies.ai - OpenAI DevDay 2026: 20+ announcements across ChatGPT, Codex and the API
OpenAI's recap gathers its DevDay launches, covering GPT-6 Astra, ChatGPT, Codex, the APIs, security and builder tools.
Source: openai.com - OpenAI launches dots, proactive assistants that keep working on projects
Dots are assistants that carry on with complex projects and everyday tasks while you stay in control.
Source: openai.com
Model 5
- Replit Agent adds GPT-6.1 Sol, Sonnet 5.5, Opus 5.5 and GPT-6 Sol
Sonnet 5.5 is in Power Mode and GPT-6.1 Sol in Max Mode; availability depends on tier and workspace policy.
Source: docs.replit.com - Claude Sonnet 5.5 in Kiro at a 1.3× credit multiplier
Experimental support for Pro and above, with a 1M-token context and adaptive thinking on by default.
Source: kiro.dev - Google announces Gemini 4 Argon, its next frontier model
Google aims Argon at real-world coding, enterprise knowledge work and cyber defense, and said it would roll out soon.
Source: blog.google - OpenAI ships GPT-6.1 Sol at one-fifth of Astra's API price
OpenAI pitches GPT-6.1 Sol as near-Astra intelligence for coding, computer use and professional work at a fifth of Astra's standard input and output token prices.
Source: openai.com - ElevenLabs releases Eleven v4 and v4 Turbo
v4 brings more expressive speech and better voice cloning in 90+ languages; v4 Turbo targets real-time agents at about 100 ms median latency.
Source: elevenlabs.io
Feature 7
- Model training now stays off for Wispr Flow Growth and Enterprise plans
The setting is tied to the plan, members can't re-enable it, and Enterprise admins can audit access to shared notes.
Source: wisprflow.ai - Runway Agent can draft with Seedance 2.5 at 480p for faster, cheaper iterations
Draft mode is suggested automatically on all plans.
Source: runway.com - Kilo Cloud Agents can share a branch and take inline feedback
Several chats can now work on the same branch at once, each keeping its own context.
Source: blog.kilo.ai - Manus 2.0 adds a Video Editor and Game Dev
Two new creation tools for making videos and games inside Manus.
Source: manus.im - Gemini adds skills, which will replace Gems
Skills are reusable instructions Gemini can build from chats and run on matching prompts; existing Gems will be migrated automatically when Gems go away.
Source: gemini.google - Canva builds a connector for Meta's Muse agent
Muse can now generate on-brand designs straight from a user's Canva account.
Source: canva.com - Amp adds Plaid speed: up to 6× faster GPT-6 Astra at 6× the token cost
Plaid uses OpenAI's ultrafast tier for GPT-6 Astra modes and works only with Amp-provided inference, not linked ChatGPT subscriptions.
Source: ampcode.com
Releases 1
- Zed 1.21–1.22: SuperGrok sign-in, Grok 4.7 default, BYOK for GPT-6 and Opus 5.5
SuperGrok subscribers can use Grok in the Agent Panel, and subagents can pick a model per spawn.
Source: zed.dev
Incident 1
- Claude.ai, Claude Code, Cowork and the API hit by errors and sign-in failures
Impact ran from 7:00 to 7:59 am PT on 29 Sep; sign-in, new chats and Code sessions failed, and some messages sent then may not have been saved.
Source: status.claude.com
Verified price & limit changes
-
new tierlimit cutChatGPT Pro 500 ($500/mo) launched; Pro 200 reopened with a lower usage allowance
Pro 500 adds Ultrafast GPT-6 Astra / GPT-6.1 Sol. New, non-grandfathered Pro 200 subscriptions get a lower allowance; grandfathered users keep the old allowance through Oct 29, 2026.