What's new
New models, features and updates on anyAInow. No subscription — top up from $5, tokens never expire.
August 2026
- New modelAug 11, 2026
Claude Opus 5, with a Fast option for when speed matters
Anthropic's newest flagship is live, and for the first time you can pick a speed variant right from the message box instead of the model picker.
July 2026
- FeatureJul 19, 2026
Parents choose exactly what a subaccount can do
New guardrail toggles for chat and study tools: give a student study tools only, chat only, or both, alongside the existing controls for images, voice, web search and models.
- FeatureJul 18, 2026
Free homework study tools for students and families
Two built-in tutors that teach instead of answering, free on any balance, plus family guardrails: model limits, a family-safe mode and rules kids can't change.
- New modelJul 17, 2026
Seven new frontier models
OpenAI's new GPT-5.6 family (Luna, Terra, Sol), Claude Sonnet 5, Grok 4.5, Kimi K3 and Qwen 3.7 Max are now in the model picker.
- RetiredJul 17, 2026
Deprecated: Mistral Large and MiMo V2 Pro
Two models left the picker, each replaced by a newer, cheaper successor: Mistral Large 3 and MiMo V2.5.
- PricingJul 17, 2026
Mistral Large 3: newer and 4x cheaper
Mistral's flagship slot now runs Large 3 - a newer, stronger model at a quarter of the old price, and it reads images too.
- FeatureJul 14, 2026
anyAInow in your language
The whole site now speaks Spanish, French, German, Italian, Dutch, Portuguese, Japanese and Chinese - switch anytime from the top-right menu.
- FeatureJul 13, 2026
Guardrails for subaccounts
Turn features on or off per member and pin instructions the AI always follows - e.g. “help with the homework but don't give answers, explain so they can learn.”
- FeatureJul 8, 2026
Continue from any compare answer
Comparing models? Pick the answer you like and hit Continue — the chat forks from that answer, on that model, so follow-ups stay on point.
- FeatureJul 8, 2026
Edit any image you've generated
A new Edit button on every generated image: describe the change and the model edits that image - no downloading and re-uploading.
- FeatureJul 8, 2026
High-quality image generation
A new Quality control in the image composer: High renders up to 2K at the model's best setting, priced by what it actually costs.
- FeatureJul 4, 2026
Voice: dictate, and a text-to-speech studio
Speak your messages with the composer mic, and turn any text or document into natural speech in the new Voice studio.
- FeatureJul 3, 2026
Connect your own tools with MCP
Add MCP servers — including OAuth ones like Sentry — and the AI can use their tools in your chats. You approve every server before it runs.
- New modelJul 2, 2026
Claude Fable 5 is here
Anthropic's newest flagship — built for long-running, autonomous coding and agent work, with a 1M-token context. Pick it in any chat.
June 2026
- FeatureJun 28, 2026
See what every message costs
Every chat and image now shows its cost — in tokens, your currency, or both. Choose what you see under Settings → Display.
- FeatureJun 28, 2026
Browse every AI image model
New pages for Nano Banana, GPT Image, FLUX, Seedream and Grok Imagine — with the per-image price and which ones can edit your photos.
- PricingJun 27, 2026
Image generation up to 70% cheaper
We moved image generation to a more cost-efficient engine and passed the savings straight to you — image models are now up to 70% cheaper, same models and quality.
- FeatureJun 27, 2026
Share a conversation with a link
Turn any chat into a public read-only link — text, interactive cards and images included. No account needed to view.
- New modelJun 25, 2026
New Models: DeepSeek R1, Mistral Large and Amazon Nova
Four new models in the picker: DeepSeek R1 reasoning, Mistral's flagship Large, and Amazon's Nova Pro and Lite.
- FeatureJun 24, 2026
Introducing Subaccounts
Share your balance with people you add, each with their own optional spend cap. No extra subscription.
- FeatureJun 23, 2026
Introducing Auto
One pick that routes every prompt to the best model — and generates an image when you ask for one.
- New modelJun 23, 2026
New Image Model: Seedream 4.5
Faithful image-to-image editing — tweak a real photo without regenerating it.
- New modelJun 23, 2026
New Image Model: Flux 2 Klein
Fast, lightweight text-to-image — quick drafts at a low token cost.
- New modelJun 23, 2026
New Image Model: Flux 2 Pro
High-fidelity text-to-image — crisp, detailed, professional results.
- FeatureJun 22, 2026
Introducing Projects
Group related chats into a project with its own instructions, default model, and file library.
- FeatureJun 22, 2026
Web Search in Chat
Tap the globe in the composer to ground answers in current information from the web.
- FeatureJun 22, 2026
Resizable Sidebar
Drag the divider to set how wide the left column is — your size is remembered.
- New modelJun 21, 2026
New Open Model: GLM 5.2
Z.ai's open-weights GLM 5.2 — strong coding & tool use, 1M context. Now in the model picker.
- New modelJun 20, 2026
New Free Model: Llama 3.3 70B
Llama 3.3 70B is now available as a free tier option.
- New modelJun 20, 2026
Llama 4 Maverick Now Available
Meta's cutting-edge reasoning model with advanced chain-of-thought capabilities.