What’s New

Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are live in the Kimi API with a 1M-token context; Kimi says full weights will be released by Nov 4, 2026

release: Codex CLI 0.164.0: /presets, inline arguments for /model and /permissions, Kitty/Sixel image previews in the terminal, enterprise MCP auth, credential masking, and the Bedrock setup wizard on by default

HeyGen Video: HeyGen Video API now makes single clips of 5 to 30 seconds for text-to-video and image-to-video, at 480p, 768p, 1080p or 2K; reference-to-video and the MCP stay limited to 15 seconds

Tencent open-sourced Hy-Nomas, an 8B interactive video world model: 720p at 24 fps, clips up to 60 s, steered by text prompts and camera paths, with world memory and LoRA fine-tuning; Apache 2.0, runs on a single RTX 4090 (INT8/FP8 with CPU offload)

integration: Incorporate your business with Stripe Atlas from inside Replit (Stripe MCP integration), with 20% off when you incorporate through Replit

Replit Agent can now create and update reusable workspace skills from your chats and projects; you approve each one before it's saved

Decisions API public beta, up to 200 questions per request → OpenAI Decisions API is generally available on gpt-6-luna, with up to 400 questions per request

availability: Claude Models on Google Cloud (Anthropic managed) is generally available, with Google Cloud IAM, billing and commitment retirement

release: OpenClaw v2026.10.6 patch release: install, update and startup fixes, Windows setup fixes, Discord command repairs, iPhone Talk fixes; now requires patched Node 24.22.1+ or 26.12.1+

Replit Monitoring Agent watches, fixes and deploys changes to your apps on its own

vector model: Ideogram's first vector model generates editable SVGs from text or turns images into vectors, in Ideogram and the API

Claude Duet launches as a research preview for Pro and Max

release: Zed v1.25 shrinks the binary by about 33% on macOS, 28% on Windows and 7% on Linux

release: OpenBot 0.45–0.47 adds live voice conversations with an agent, and dictation plus read-aloud in 31 languages (Live Voice in desktop app announced Oct 22)

Share a Replit Agent conversation with teammates in your workspace

prompt caching: OpenAI API adds a 12-hour prompt caching option; 12-hour cache writes are billed at 2.5x the input price

Locked chats in ChatGPT for Free, Go, Plus and Pro, rolling out gradually

ChatGPT Slides get real-time collaboration for Pro, Business and Enterprise

Sites limit fixed number of Sites → ChatGPT Sites no longer has a fixed limit on the number of Sites for Plus and Pro

Voxtral 3: Mistral releases Voxtral 3 in public preview: Voxtral TTS 3, Mini Transcribe 3, Mini Transcribe Realtime 3 and Small Transcribe 3, plus Voice Agents

Krea 3: Krea 3 is out of beta and Krea 3 Ultra is available in the Krea API

API pricing: MiniMax M3.1 Flash launched in the MiniMax API at $0.15 input / $0.60 output per 1M tokens (cache read $0.015)

Eleven Realtime: ElevenLabs launches Eleven Realtime (v1 Mini as default, v1 Max) and Realtime Translation for enterprise workspaces

Kiro adds Memory, carrying useful context across chat sessions (Memory setting defaults to Off since Oct 20)

Slack MCP integration in v0 on all plans: use Slack discussions as context and post updates to Slack

release: OpenClaw v2026.10.5: agents sidebar, Skill Workshop 2.0 with Undo, public conversation links without sign-in

release: OpenBot 0.44: agents ask questions with buttons and forms, a new Tasks page, a yearly activity recap, and agent memory limits on joined servers

models DeepSeek V4.1 Flash → DeepSeek V4.1 Flash upgraded to DeepSeek-V4.1-Flash-1017 on the web, app and API (model name deepseek-flash, 1M context, 384K max output); API calling method and pricing unchanged. Legacy names deepseek-v4-flash and deepseek-v4-flash-vision-exp are retired and now served by the new model at the Flash price

API pricing: Search and Fetch APIs (DeepSeek's in-house search engine) open to everyone at $0.75 per 1,000 successful requests each; server-side search/fetch calls made through DeepSeek Harness, Claude Code or Codex are free until per-request billing starts at 00:00 Beijing time on 2026-11-01

Usage multipliers: Premium 4x, Super 15x, Ultra 50x Plus

Bundle: Sold as part of The X Subscription; pool shared with Grok

Teams Cursor Token Rate: $0.25 per million tokens on third-party models

Free grants: 5 daily build credits (up to 30/mo) + 20 Cloud credits/mo

Credit expiry: Monthly credits expire 2 months after issue; top-ups 12 months

Free tokens: 300K/day, 1M/month

Starter: Daily Agent credits up to monthly cap; monthly cloud credits

Pro rollover: Unused credits roll over two months

Terminology: GitHub AI Credits; premium requests = legacy

Pricing URL: windsurf.com/pricing redirects to devin.ai/pricing

Prices: Free $0, Pro $20, Max $200, Teams $80/mo min + $40/full seat

Quota model: Pro daily+weekly; Max weekly no daily cap; on-demand credits past quota

Plus price confirmed: $20/mo billed monthly (help center)

Pro Ultrafast usage rate: 8x Standard (included) / 6x Standard credit rate (Pro 500 only)

Pro Fast mode usage rate: 2.5x Standard (included allowance)

Business Premium seat usage: 5x Standard seat usage, no five-hour limit

Premium conflict: docs.x.ai Grok Bot FAQ says Grok Bot is included with every paid individual Cursor plan; Cursor help (plans page) says on The X Subscription Plus does not include it (only Premium/Super/Ultra). FAQ appears to describe legacy plans.

Super conflict: docs.x.ai Grok Bot FAQ says Grok Bot is included with every paid individual Cursor plan; Cursor help (plans page) says on The X Subscription Plus does not include it (only Premium/Super/Ultra). FAQ appears to describe legacy plans.

Ultra conflict: docs.x.ai Grok Bot FAQ says Grok Bot is included with every paid individual Cursor plan; Cursor help (plans page) says on The X Subscription Plus does not include it (only Premium/Super/Ultra). FAQ appears to describe legacy plans.

Pro Plan mode cost: 1 credit per message

Pro monthly credits: 1,000

Pro+ monthly credits: 2,000

Pro Max monthly credits: 5,000

Power monthly credits: 10,000

Pro conflict: factory.ai/pricing lists Droid Computers under Plus; Factory docs list Managed Droid Computers under Pro.

Pro rate-limit windows: 5-hour, 7-day and 30-day rolling

Plus conflict: factory.ai/pricing lists Droid Computers under Plus; Factory docs list Managed Droid Computers under Pro.

Teams conflict: devin.ai/pricing reads '$80/month for team plan + $40/mo per full dev seat'; Devin docs say $80/month is a minimum that can be met by full seats ($40 each) and on-demand credits (e.g. 2 full seats = $80 total, N>=2 seats = N x $40).

Starter conflict: Base44 Credits doc says to get more message credits you upgrade the plan; Billing doc mentions buying credit top-ups. Overage left not_confirmed.

Free daily cap: 5 message credits/day (Free)

Everyday Usage plan added: $5/mo; Muse Spark, Muse Wonder, Muse Image; 10-50 prompts every 5 hours (incl. image/video uploads); voice mode; web search

High Usage plan added: $15/mo; Everything in Everyday; 5x Everyday usage; more multimodal inputs

Power Usage plan added: $50/mo; Everything in High; Muse Wonder Full; 20x Everyday usage; early access; higher file uploads

Maximum Usage plan added: $200/mo; Everything in Power; 80x Everyday usage

Free plan added: $0/mo; included usage Limited

Plus plan added: $8/mo; included usage Base

Premium plan added: $30/mo; included usage 4x Plus

Super plan added: $100/mo; included usage 15x Plus

Ultra plan added: $200/mo; included usage 50x Plus

Free plan added: $0/mo; included usage Limited

Plus plan added: $8/mo; included usage Base

Premium plan added: $30/mo; included usage 4x Plus

Super plan added: $100/mo; included usage 15x Plus

Ultra plan added: $200/mo; included usage 50x Plus

Pro 200 allowance: New non-grandfathered Pro 200 subscriptions get lower allowance; grandfathered through 2026-10-29

Claude lineup: Fable 5.1, Opus 5.5, Sonnet 5.5, Haiku 5.5 current; Sonnet 5 $2/$10 made permanent (planned $3/$15 rise cancelled)

Cloud agents can build interactive canvases and share them by link; Environments API in beta; cloud agent environments support up to 300 repositories and large ones load 5x faster

GPT-6 Pro in the chat tab: faster responses, better plugin and tool use, and multiple agents working on parts of a task in parallel; rolling out to eligible Pro users

Core and Pro can choose Claude Haiku 5.5 in the model selector in Power Mode

Bring Your Own Cloud for Enterprise: Full BYOC (private preview), BYOC for Published Apps (early access) and Image Exports (early access) on Azure, Google Cloud or AWS

API pricing: Agent API supports google/gemini-4-argon; introductory pricing through 2026-11-30: $2.00/M input, $0.10/M cached input, $10.00/M output (reasoning tokens included)

Cline Cloud available to every Cline user: tasks run in their own cloud sandbox and keep going with the laptop closed; up to 10 sessions at once

Junie Local runs on NVIDIA RTX Spark from day one and Windows support is stable; free local coding agent with no tokens or quotas and no code leaving the machine

Muse Wonder in Cursor at $4/$20 per M, cache read $0.20

Kimi K3 (Moonshot AI, 1M-token context, text and images) available with experimental support in Kiro IDE, CLI, Crew and Web for Pro, Pro+, Pro Max and Power customers using the US geography

GPT-5.5 retired from Codex with ChatGPT sign-in

Gemini 4 Argon rolling out to Google AI Pro and Ultra in the Gemini app; API gemini-4-argon $2/$10 per M through 2026-11-30, $4/$20 from 2026-12-01; not on API Free tier or Google AI Pro plan (API page)

Gemini 4 Argon in Cursor at $2/$10 per M (third-party)

X Subscription: Cursor joins The X Subscription (Free/Plus $8/Premium $30/Super $100/Ultra $200); one weekly pool with Grok; third-party models move to API pricing on new plans; legacy plans continue

X Subscription: Grok part of The X Subscription from $8/mo; SuperGrok users can switch