The AI map · Models and assistants

Frontier models

The most capable models from the big labs, sold in tiers: a top model for the hardest work, an everyday model, and a cheap, fast one for volume. The ones that lead it as of 29 October 2026.

  1. GPT-6 family: GPT-6 Astra, GPT-6.1 Sol, GPT-6.1 Terra, GPT-6 Luna New since July

    OpenAI

    Astra for the hardest reasoning, coding and computer use; 6.1 Sol for everyday complex work near Astra's level; 6.1 Terra for fast visual reasoning, computer use and writing; Luna for cheap, high-volume tasks.

    Pick it You want one family that covers every tier; Sol is the usual default, Astra for the hardest jobs, Luna when cost per call matters most.

    Price API per million tokens (input / output): Astra $10 / $50; 6.1 Sol $2 / $10; 6.1 Terra $1 / $5; Luna $0.10 / $0.50

    Source: openai.com

  2. Claude Fable 5.1 New since July

    Anthropic

    Anthropic's top public model, for demanding reasoning and long agent runs; the same model as Mythos 5.1 with cyber and biology safeguards.

    Pick it Opus 5.5 at higher effort still falls short on a hard, long task.

    Price API $10 per million input tokens, $50 per million output

    Source: platform.claude.com

  3. Claude 5.5 family: Opus 5.5, Sonnet 5.5, Haiku 5.5 New since July

    Anthropic

    Opus 5.5 is Anthropic's recommended starting point for coding and knowledge work; Sonnet 5.5 balances speed and intelligence; Haiku 5.5 is for high-volume work like classification, extraction and routing.

    Pick it You build long agentic coding or writing work; Anthropic says Opus 5.5 performs at Fable 5.1's level on most work for less.

    Price API per million tokens (input / output): Opus 5.5 $4 / $20; Sonnet 5.5 $2 / $10; Haiku 5.5 from $0.10 / $0.50

    Source: platform.claude.com

  4. Gemini 4 Argon New since July

    Google

    Google's flagship reasoning model for complex software, maths, science and multi-step agents, with long output up to 1M tokens.

    Pick it You want a top-tier model at a lower API price while Google's introductory rate lasts.

    Price API $2 per million input, $10 per million output until 30 November 2026; $4 and $20 from 1 December 2026

    Source: ai.google.dev

  5. Gemini 3.8 Flash New since July

    Google

    Google's most capable Flash model, an everyday workhorse for coding, agents and enterprise work, with a free API tier.

    Pick it You want a fast, cheap model with a free tier to start building on.

    Price Free tier; paid API $0.75 per million input, $3.75 per million output until 31 December 2026, then $1.50 and $7.50

    Source: ai.google.dev

  6. Grok 4.7 New since July

    xAI (SpaceXSI)

    xAI's most capable model, for coding and knowledge work.

    Pick it You want a capable general model with cheap output tokens, or already use Grok.

    Price API $2 per million input, $6 per million output

    Source: docs.x.ai

  7. Muse Spark 1.1 New since July

    Meta (Meta Superintelligence Labs)

    Meta's multimodal reasoning model for agents, computer use and coding, behind Meta AI and a new Meta Model API.

    Pick it You want to try Meta's own frontier model; the API is in public preview.

    Source: ai.meta.com

What changed

  1. 26 Oct 2026 OpenAI released GPT-6.1 Terra at $1 / $5 per million tokens, half the price of GPT-6.1 Sol. openai.com
  2. 13 Oct 2026 Google made Gemini 4 Argon, its new flagship, generally available. ai.google.dev
  3. 8 Oct 2026 Google deprecated Gemini 3.7 Flash and 3.5 Flash in favour of 3.8 Flash and 3.6 Flash. ai.google.dev
  4. 7 Oct 2026 Anthropic released Claude Haiku 5.5, which it says costs around 75% less to run than Haiku 4.5. anthropic.com
  5. 29 Sep 2026 OpenAI released GPT-6.1 Sol, which it says nearly matches Astra at a fifth of the price, and Ultrafast, a faster speed tier. openai.com
  6. 28 Sep 2026 Anthropic released Claude Sonnet 5.5; OpenAI said it would not release GPT-6.1 Astra because it did not meet its safety standards. cnbc.com
  7. 22 Sep 2026 Anthropic released Claude Opus 5.5; the same day OpenAI released GPT-6 Sol and GPT-6 Luna at half the price of their GPT-5.6 versions. openai.com
  8. 21 Sep 2026 xAI released Grok 4.7. x.ai
  9. 3 Sep 2026 OpenAI released GPT-6 Astra, its most capable model, at $10 / $50 per million tokens in the API. x.com
  10. 2 Sep 2026 Google made Gemini 3.8 Flash generally available. ai.google.dev
  11. 1 Sep 2026 Anthropic released Claude Mythos 5.1 for vetted organisations, with Claude Fable 5.1 as the same model with safeguards for everyone else. anthropic.com
  12. 13 Aug 2026 Google made Gemini 3.7 Flash generally available. ai.google.dev
  13. 21 Jul 2026 Google made Gemini 3.6 Flash and Gemini 3.5 Flash-Lite generally available. ai.google.dev
  14. 16 Jul 2026 xAI released Grok 4.5, followed by Grok 4.6 on 12 August. x.ai
  15. 9 Jul 2026 Meta released Muse Spark 1.1 with a public preview of the Meta Model API. ai.meta.com

Questions people ask about it

AI in Your Product

  1. Should my startup add AI features, and which ones actually matter to users?
  2. How do founders ship their first AI feature (APIs, wrappers, RAG in plain terms)?
  3. How much does it cost to run AI features, and how do I keep costs sane?
  4. Which AI model should I build on (GPT, Claude, Gemini, open-source), and how do I avoid locking myself in?
  5. How do I test whether my AI feature actually works before launch (evals, in plain terms)?
  6. How do I stop my AI feature from making things up in front of customers?

Prompting

  1. Why do some people get 10x better results from the same AI tools?
  2. What are the 5 prompting techniques worth learning (and the ones that are outdated)?
  3. How do I give AI the right context about my business (projects, custom instructions, memory)?

Building Your AI Stack

  1. ChatGPT vs Claude vs Gemini vs Grok, how do founders actually choose?
  2. Which paid AI subscriptions are worth it for an early-stage founder?
  3. How do I evaluate the endless stream of new AI tools without drowning?

The rest of the map

Checked 29 October 2026. Prices and rankings move every few months; each tool links the page its claims were read from.

eChai Partner Brands The partnership →