Agentic WeeklyGet the weekly

AI model questions answered: pricing, release dates and head-to-heads

Short answers to what people ask AI about AI models, each with a date and a source. Where a model is not in our data (Grok 4.8, Llama 5, Gemini 4 Carbon, a DeepSeek V4 Pro) we say so instead of guessing. Prices and dates are read from vendor pages and our model records, last checked Oct 27, 2026. Open a question to read the answer.

By Agentic Weekly, operated independently from Texas, US · Last updated (America/Chicago) · How we work

Free weekly brief: email newsletter · RSS

Evidence: Reported. Facts come from our model records (vendor pages read Oct 26, 2026) and from chatgpt.com/pricing read Oct 27, 2026 (America/Chicago). We did not run these models. Where we found nothing, the answer says so.

How to read these answers

  • Each answer starts with the answer, then the date and source.
  • "Not in our data" means we found no vendor announcement; it does not mean the model will never exist.
  • Rumored and expected releases are tracked on upcoming models, marked Unconfirmed until a vendor says so.

FAQ

What are ChatGPT's pricing plans?

ChatGPT has six plans: Free, Go, Plus, Pro, Business and Enterprise (chatgpt.com/pricing, read Oct 27, 2026). The page text we read shows no dollar amounts, so we quote none. Codex is limited on Free and Go, included on Plus and expanded on Pro. Plus includes GPT-6.1 Sol, GPT-6 Astra and GPT-6 Sol. See ChatGPT notes.

What is Gemini 4 Carbon, or Gemini 4 Pro?

Not in our data. Google's pricing page lists Gemini 4 Argon and no model named Carbon, and our data has no model called Gemini 4 Pro. Gemini 4 Argon was released Oct 13, 2026 at $2 / $10 per 1M tokens, rising to $4 / $20 from Dec 1, 2026 (see Argon pricing). Rumors are on upcoming models.

When is Grok 4.8 coming out?

We found no xAI announcement of Grok 4.8, so we have no date. The newest Grok in our data is Grok 4.7, on the API since Sep 21, 2026. See Grok 5 release date for what xAI has and has not said (read Oct 27, 2026).

Grok 4.7 vs Claude Opus 5.5: which is better?

We have not benchmarked them. On price and size: Grok 4.7 is $2 / $6 per 1M tokens with a 0.5M-token window; Claude Opus 5.5 is $4 / $20 with a 1M-token window. Opus 5.5 is #1 on LMArena WebDev (Oct 22, 2026); Grok 4.7 is not among the rows we cite.

DeepSeek V4 Flash vs Pro: what is the difference?

Our data has DeepSeek-V4.1-Flash (Sep 10, 2026) and its 1017 update (Oct 17, 2026), MIT-licensed, with a 1M-token window and prices of CNY 1 input / CNY 4 output per 1M tokens (peak hours CNY 2 / 8). We have no record of a V4 Pro; DeepSeek's page announces none. See DeepSeek-V4.1-Flash-1017.

Kimi K3 vs Claude Fable 5: how do they compare?

On list data: Kimi K3 is open-weights (custom Kimi K3 license) at $3 / $15 per 1M tokens with a 1M-token window; Claude Fable 5 is proprietary at $10 / $50 with a 1M-token window. On LMArena WebDev (Oct 22, 2026) Claude Fable 5.1 is #5 and Kimi K3 #17; Fable 5 itself is not listed. Preference votes, not our tests.

When is Llama 5 released?

We found no Meta announcement of Llama 5, so no date; it is listed as Unconfirmed on upcoming models. The newest Meta open-weights model in our data is Muse Spark Open, released Oct 15, 2026 under the Muse Model License (limited commercial terms); see its licence guide.

Sources

Page history

  • : First published. Facts read from the sources listed below on this date.

Markdown version for agents: /what-is/ai-model-questions-answered.md