Skip to content

Anthropic

Dossier checked 28 Sep 2026

Claude on the free plan Claude Sonnet 5.5 is on the free plan. Claude Opus 5.5 and Claude Fable 5.1 are not. Claude Mythos 5.1 is restricted to vetted access. The Access Gap, checked 3 Oct 2026.

What it sells now

Newest frontier release

Claude Sonnet 5.5

28 Sep 2026 · the lab’s announcement

Mid-tier model at $2 in and $10 out, on every platform from launch

Flagship list price, per million tokens

$10 in $50 out

Claude Fable 5.1, the most capable generally available model, released 1 Sep 2026 · pricing page

Fable 5.1 is the top price tier. Anthropic calls Claude Opus 5.5 (22 Sep) its “new leading model”, at Fable 5.1’s level on most work, and Sonnet 5.5 (28 Sep) is its mid tier.

Weights

Closed

None. Every Claude model is closed; access is through the API, the apps or a cloud partner. Source

Where you can buy the flagship

  • Claude API
  • Amazon Bedrock
  • Google Cloud (Vertex AI)
  • Microsoft Foundry
  • Claude Platform on AWS
  • OpenRouter
  • claude.ai, the apps and Claude Code

Source

14 releases in the year to 30 Oct 2026, 6 the year before · 27 in all since 2023

Our read, not a fact

In what others build on

The question we ask of every lab: is most of its value in the model, which gets cheaper fastest, or in something other work is built on? Our answer: in something other work is built on.

Anthropic is climbing the stack toward the layer where models do real work. Claude Code and the agent tooling are its durable position; the raw model underneath commoditises like everyone else’s.

What would change our readA rival agent tool matches Claude Code’s reliability, or Anthropic’s model quality falls far enough that the tooling can no longer carry it.

01 · What it runs on, and the money

Trains and serves on

Claude trains and runs on AWS Trainium, Google TPUs, NVIDIA GPUs and AMD Instinct. AWS is the named primary training and cloud provider, under a commitment of more than $100B over ten years. AMD agreed in July 2026 to supply up to 2 GW of MI450 GPUs, building on Anthropic’s existing use of AMD chips, and in May 2026 Anthropic secured GPU capacity at SpaceX’s Colossus sites. Source 1 Source 2 Source 3

What it depends on Four chip families are contracted and AWS is named primary. Amazon, Google, NVIDIA and AMD each supply chips and have invested or committed to invest, so its suppliers and its shareholders overlap. Source

Biggest risk More than $100B is committed to AWS over ten years and AWS is the named primary provider, while Amazon is also one of the largest shareholders. Source

Raised
Series H $65B (May 2026), led by Altimeter, Dragoneer, Greenoaks and Sequoia Source
Valuation
$965B post-money (28 May 2026) Source
Revenue
Run rate above $47B (May 2026) Source
Structure
A Delaware public benefit corporation. A Long-Term Benefit Trust, whose trustees hold no equity, holds a share class that elects part of the board. It confidentially submitted a draft IPO registration to the SEC on 1 Jun 2026. Source

NotesPer-token prices understate the cost of text since Claude Opus 4.7, whose tokenizer produces roughly 30% more tokens for the same input. Press reports put the run rate above $65B at the end of July (Bloomberg, Aug 2026) and describe a proposal to give the founders majority voting control (The Information, Sep 2026); Anthropic has confirmed neither. Headcount and total funding are not published.

02 · Releases

data through 30 Oct 2026

  1. Anthropic 21 Claude Sonnet 5.528 Sep 2026

Each dot is a frontier release at its date. Click one to find it in the full log. Every lab on one chart · Which plans and apps its new models reached

The landmarks

  1. 24 Jul 2026 Claude Opus 5 Opus-line generational: near-Fable-5 intelligence at half the price Flagship
  2. 30 Jun 2026 Claude Sonnet 5 Sonnet-line generational: close to Opus 4.8 at a fraction of the price Generational
  3. Jun 2026 Claude Fable 5 First generally available Mythos-class model, with safety classifiers; 1M context. Mythos 5, the same model without them, stayed partner-only Mythos-class1M ctx
  4. May 2026 Claude Opus 4.8 Flagship of the 4.x line Flagship
  5. Apr 2026 Claude Mythos Preview First Mythos-class model, a tier above Opus. Never generally released: critical-infrastructure partners only Mythos-classRestricted
  6. Nov 2025 Claude Opus 4.5 Flagship reasoning, frontier coding ReasoningFlagship
  7. May 2025 Claude Opus 4 / Sonnet 4 Claude 4 generation, stronger agents AgenticFlagship
  8. Feb 2025 Claude Code Agentic coding tool in the terminal AgenticCoding CLI
  9. Feb 2025 Claude 3.7 Sonnet First hybrid reasoning model ReasoningHybrid
  10. Jun 2024 Claude 3.5 Sonnet Major coding leap, Artifacts launch CodingFlagship
  11. Mar 2024 Claude 3 (Opus/Sonnet/Haiku) Three-tier family, first multimodal Claude MultimodalThree tiers
  12. Mar 2023 Claude 1 First Claude, Constitutional AI approach; the API was by request, the free Claude app for Slack open to any team from 30 Mar First Claude
The full release logall 27, newest first
  1. 2026
  2. 7 Oct 2026 Claude Haiku 5.5 Fast tier refresh, tuned for high-volume and latency-sensitive work Source
  3. 28 Sep 2026 Claude Sonnet 5.5 Frontier Mid-tier model at $2 in and $10 out, on every platform from launch Source
  4. 22 Sep 2026 Claude Opus 5.5 Frontier Available on every platform from launch Source
  5. 1 Sep 2026 Claude Mythos 5.1 Restricted sibling of Fable 5.1 for vetted cyberdefence and life-science organisations Source
  6. 1 Sep 2026 Claude Fable 5.1 Frontier Generally available; its sibling Mythos 5.1 stayed partner-only Source
  7. 24 Jul 2026 Claude Opus 5 Frontier Opus-line generational: near-Fable-5 intelligence at half the price Source
  8. 30 Jun 2026 Claude Sonnet 5 Frontier Sonnet-line generational: close to Opus 4.8 at a fraction of the price Source
  9. Jun 2026 Claude Fable 5 Frontier First generally available Mythos-class model, with safety classifiers; 1M context. Mythos 5, the same model without them, stayed partner-only
  10. May 2026 Claude Opus 4.8 Frontier Flagship of the 4.x line
  11. Apr 2026 Claude Opus 4.7 Frontier Long-horizon agentic coding
  12. Apr 2026 Claude Mythos Preview First Mythos-class model, a tier above Opus. Never generally released: critical-infrastructure partners only
  13. 17 Feb 2026 Claude Sonnet 4.6 Frontier Balanced tier upgrade; the default for Free and Pro on claude.ai, and on the API, from launch Source
  14. Feb 2026 Claude Opus 4.6 Frontier Agent-team orchestration
  15. 2025
  16. Nov 2025 Claude Opus 4.5 Frontier Flagship reasoning, frontier coding
  17. Oct 2025 Claude Haiku 4.5 Fast tier refresh
  18. 29 Sep 2025 Claude Sonnet 4.5 Frontier Best agent value tier
  19. Aug 2025 Claude Opus 4.1 Frontier Coding and agent capability upgrade
  20. May 2025 Claude Opus 4 / Sonnet 4 Frontier Claude 4 generation, stronger agents
  21. Feb 2025 Claude Code Agentic coding tool in the terminal
  22. Feb 2025 Claude 3.7 Sonnet Frontier First hybrid reasoning model
  23. 2024
  24. Oct 2024 Claude 3.5 Sonnet (upgraded) Frontier Upgraded Sonnet, shipped with computer use in beta
  25. Oct 2024 Claude 3.5 Haiku Fast, low-cost tier refresh
  26. Jun 2024 Claude 3.5 Sonnet Frontier Major coding leap, Artifacts launch
  27. Mar 2024 Claude 3 (Opus/Sonnet/Haiku) Frontier Three-tier family, first multimodal Claude
  28. 2023
  29. Nov 2023 Claude 2.1 Frontier Doubled context, fewer refusals
  30. Jul 2023 Claude 2 Frontier First public release, long context: claude.ai open to anyone in the US and UK
  31. Mar 2023 Claude 1 Frontier First Claude, Constitutional AI approach; the API was by request, the free Claude app for Slack open to any team from 30 Mar

03 · When to reach for them

practical read

Agentic coding
Reach for Claude Code when the work is multi-step engineering in a real repository (Feb 2025).
Long-horizon work
Pick Opus 5.5 for long-running agentic coding and knowledge work, and Fable 5.1 when that falls short (Sep 2026).
Safety-sensitive
Use it where the provider’s safety commitments are part of the decision; its most capable model, Mythos 5.1, stays with vetted programmes (Sep 2026).

04 · Weights, across all ten

the flagship

Open flagship
DeepSeek Mistral AI Alibaba (Qwen) Moonshot AI Zhipu (Z.ai)
Closed flagship, some open
OpenAI Google Meta xAI
Closed
Anthropic

05 · Essays that read Anthropic

from the archive

The Model Is Not the Moat If frontier capability keeps centralising, the durable edge shifts outward into trust, workflow fit, and the surrounding package. The Engine Underneath Hard Decisions Eight stages turn hidden structure into durable knowledge. Most teams run three of them and call it understanding. The other five are where compounding hides. The Five Laws of Durable Systems What still has a job after the change? Five tests for seeing what is likely to survive.

06 · About Anthropic

at a glance

Founded
2021
HQ
San Francisco, USA
Leadership
Dario & Daniela Amodei
Backers
Amazon, Google, Microsoft, NVIDIA
Also builds
Claude Code, Claude apps; created MCP
Known for
Claude, AI safety

Founded in 2021 by former OpenAI researchers including siblings Dario Amodei (CEO) and Daniela Amodei. Headquartered in San Francisco, it builds the Claude family with a safety-first "Constitutional AI" approach, offered as closed-weight models, and its agentic coding tool, Claude Code.