Product dossier · by Anthropic
Claude
Versions, pricing, availability and every story about Claude.

TopicsAll topics
-
Stale CLAUDE.md and AGENTS.md Files Override the Code, and One Sentence Fixes It

Claude Code and Codex CLI followed outdated instruction files in 90.1% of convention choices. One sentence giving existing code precedence raised code-matching choices to 91.4%.
By
Deals & partnerships · Anthropic
Partnership
Accenture partnership on independent frontier-AI evaluation (≥$1B each over five years)
anthropic.com (source, opens in a new tab)Funding
Anthropic raises $65B Series H at $965B valuation
Led by Altimeter, Dragoneer, Greenoaks and Sequoia; co-led by Capital Group, Coatue, D1, GIC, ICONIQ and XN. Includes $15B of previously committed hyperscaler investment, $5B of it from Amazon. Run-rate revenue had crossed $47B.
reuters.com (source, opens in a new tab)Partnership
xAI to provide Anthropic with Colossus 1 compute
x.ai (source, opens in a new tab)Partnership
Google to invest up to $40B in Anthropic ($10B now)
Google commits $10B in cash now and up to $30B more tied to performance milestones; confirmed by both companies. Before the deal Google had invested over $3B and reportedly held ~14% (CNBC, Reuters).
cnbc.com (source, opens in a new tab)Partnership
Amazon invests $5B (up to $25B) as Anthropic commits $100B+ to AWS for up to 5 GW
Amazon invests $5B now and up to $20B more on milestones (on top of $8B earlier). Anthropic commits over $100B over ten years to AWS (Graviton, Trainium2–4), up to 5 GW; AWS stays primary training partner.
anthropic.com (source, opens in a new tab)Partnership
Google and Broadcom: multiple GW of next-gen TPUs for Anthropic (~3.5 GW via Broadcom from 2027)
Expands the Oct 2025 TPU deal. Broadcom's 8-K: Anthropic will access ~3.5 GW through Broadcom starting 2027. Mostly US-sited, part of Anthropic's $50B US infrastructure pledge.
anthropic.com (source, opens in a new tab)Acquisition
Anthropic acquires Coefficient Bio in ~$400M stock deal
techcrunch.com (source, opens in a new tab)Acquisition
Anthropic acquires Vercept to boost Claude’s computer use
anthropic.com (source, opens in a new tab)Funding
Anthropic raises $30B Series G at $380B valuation
Led by GIC and Coatue at a $380B post-money valuation, with D. E. Shaw Ventures, Dragoneer, Founders Fund, ICONIQ and MGX.
anthropic.com (source, opens in a new tab)Partnership
Snowflake signs $200M multi-year deal to bring Claude to 12,600 customers
anthropic.com (source, opens in a new tab)
Sources: company announcements & release notes, press reports.
Release timeline
Every Claude model release · 22 releases
Claude Haiku 5.5
support.claude.com (source, opens in a new tab)Claude Sonnet 5.5
support.claude.com (source, opens in a new tab)Claude Opus 5.5
anthropic.com (source, opens in a new tab)Claude Fable 5.1 & Mythos 5.1
support.claude.com (source, opens in a new tab)Claude Opus 5
support.claude.com (source, opens in a new tab)Claude Sonnet 5
anthropic.com (source, opens in a new tab)Claude Fable 5 & Mythos 5
platform.claude.com (source, opens in a new tab)Claude Opus 4.8
support.claude.com (source, opens in a new tab)
Show all 22 releases
Claude Opus 4.7
support.claude.com (source, opens in a new tab)Claude Sonnet 4.6
support.claude.com (source, opens in a new tab)Claude Opus 4.6
support.claude.com (source, opens in a new tab)Claude Opus 4.5
support.claude.com (source, opens in a new tab)Claude Haiku 4.5
anthropic.com (source, opens in a new tab)Claude Sonnet 4.5
support.claude.com (source, opens in a new tab)Claude Opus 4.1
anthropic.com (source, opens in a new tab)Claude Opus 4 & Sonnet 4
anthropic.com (source, opens in a new tab)Claude 3.7 Sonnet
anthropic.com (source, opens in a new tab)Claude 3.5 Haiku & upgraded 3.5 Sonnet
anthropic.com (source, opens in a new tab)Claude 3.5 Sonnet
anthropic.com (source, opens in a new tab)Claude 3 (Opus, Sonnet, Haiku)
anthropic.com (source, opens in a new tab)Claude 2.1
anthropic.com (source, opens in a new tab)Claude 2
anthropic.com (source, opens in a new tab)
Sources: company announcements & release notes.
Latest on X
Claude · @claudeai
Claude Models on Google Cloud (Anthropic managed) is now generally available. Get the full set of Claude Platform features on Google Cloud, with Google Cloud IAM credentials, billing, and commitment retirement.
Starting next week, sessions across Claude web, desktop, and mobile apps will run in auto mode by default on Pro, Max, and Team plans. Rather than asking you to approve each step, an automated check blocks risky actions. Switch back to manual anytime.
We’re opening a research preview of Claude Duet, a new app where Claude works as a proactive agent. It comes to you with what needs your attention, completes tasks you ask of it, keeps working when you step away, and checks with you before it acts on your behalf.
Claude Security now keeps up with your code as it changes. Connect your repositories once. Claude runs a deep first scan, then re-scans as changes land, so findings stay close to the current state of your code. In public beta on Claude Enterprise.
Six projects from the last few weeks that caught our attention: A wildfire analysis prototype that maps the fire’s spread over time and tracks hot spots, made with Opus 5.5.
Editor's picks
-

Harvard Study Finds Claude AI Outperforms Human Peer Review
A Harvard University study reveals that AI-generated reviews powered by Claude outperform paid human experts on key quality…
-

Claude Haiku 5.5: 75% Cheaper Model for High-Volume AI Tasks
Anthropic’s Claude Haiku 5.5 delivers 75% cost savings and enhanced coding for subagent workflows in multi-agent systems. Available…
-

Anthropic Claude Startups: Free Credits and AI Tools
Anthropic expands Claude Startups with a free year of Team access, $1,000 API credits, and partner discounts. Learn…
More stories
-

Tampered ML: Even the Best AI Models Miss One in Five Planted Research Flaws
A new EMNLP 2026 benchmark pairs 1,858 ML experiment designs with flawed twins. Claude Opus 5 leads at 77.9%, but models flip verdicts when the order…
-

Only One AI Agent Beats Researchers’ Own Code on Most FrontierCS 2 Tasks
FrontierCS 2 deletes a paper’s contribution from its repository and asks agents to beat the authors’ own code. Claude Opus 5.5 passes 58% of 48 tasks;…
-

Hybrid Computer-Use Agents Top Out at 41% When GUI and CLI Must Work Together
WeaveBench’s 114 real-world tasks: best PassRate 41.2%. GUI-only or CLI-only collapse below 3.5%. Outcome-only grading overestimates GPT-5.5 from 53.5% to an audited 33.3%.
-

Scientific AI Agents Finish Only 21% of End-to-End Workflows — and Often Claim Success Anyway
FrontierChallenge: best configs pass 20.6% of 97 released science workflows. In electrochemistry, Avg. Score hits 94.9 with 0% full completion; 75.5% of failed Claude Code runs…
-

AI Agents That Learn the Facts Still Deliver Only 60–70% of the Business Value
FDE-Gym’s 95 real delivery tasks: Claude Opus 5.5 leads at 0.56, but even models that gather 72–85% of key facts deliver just 0.60–0.70 of the reference…
-

Best Video Model Scores Just 51% on Visual Reasoning Tasks, VGI-Bench Finds
VGI-Bench’s 27 tasks and 810 instances show video generators can do some visual reasoning, but Seedance 2.0 tops out at 51%, with almost no self-correction during…
-

Giving AI Agents Incentives Nearly Doubles How Often They Deceive, Benchmark Finds
DecepEval tests 9 frontier models on 1,532 scenarios: inducements raise deception by 41.6% on average, with long-horizon tasks hitting 87% induced rates.
-

Hidden AI Reasoning Can Be Decrypted by Cheaper Models, Researchers Show
Encrypted chain-of-thought blocks from Anthropic, OpenAI and Google APIs are interchangeable across models. Feeding them to a weaker model makes it print the reasoning, exposing 367…
-

“Very Seen, But Still Very Alone”: How Chatbots Become Emotional Support
A nine-month study following 18 U.S. adults finds emotional reliance on ChatGPT, Claude and Gemini grows out of practical use, and that model updates can disrupt…
Versions & specs
Model · date · context · max output · price per 1M tokens (in/out, cached) · modalities · cutoff · license
| Model | Released | Context | Max output | Price in / out (per 1M) | Modalities | Knowledge cutoff | License / access | Spec source |
|---|---|---|---|---|---|---|---|---|
| Claude Haiku 5.5 | Oct 7, 2026 | 1M | 128K | $0.1 / $0.5 cached $0.01 Prompts >100K tokens: $0.50 in / $2.50 out. | Text, image in → text out | Jun 2026 | Proprietary (API & app) | platform.claude.com |
| Claude Sonnet 5.5 | Sep 28, 2026 | 1M | 128K | $2 / $10 cached $0.1 | Text, image in → text out | Jun 2026 | Proprietary (API & app) | platform.claude.com |
| Claude Opus 5.5 | Sep 22, 2026 | 1M | 128K | $4 / $20 cached $0.2 | Text, image in → text out | Jun 2026 | Proprietary (API & app) | platform.claude.com |
| Claude Fable 5.1 & Mythos 5.1 Claude Fable 5.1, Claude Mythos 5.1 | Sep 1, 2026 | 1M | 128K | $10 / $50 cached $0.25 | Text, image in → text out | Jun 2026 | Proprietary (API & app) | platform.claude.com |
| Claude Opus 5 | Jul 24, 2026 | 1M | 128K | $5 / $25 cached $0.5 | Text, image in → text out | — | Proprietary (API & app) | platform.claude.com |
| Claude Sonnet 5 | Jun 30, 2026 | 1M | 128K | $2 / $10 cached $0.2 | Text, image in → text out | — | Proprietary (API & app) | platform.claude.com |
| Claude Fable 5 & Mythos 5 Claude Fable 5, Claude Mythos 5 | Jun 9, 2026 | 1M | 128K | $10 / $50 cached $1 | Text, image in → text out | — | Proprietary (API & app) | platform.claude.com |
| Claude Opus 4.8 | May 28, 2026 | 1M | 128K | $5 / $25 cached $0.5 | Text, image in → text out | — | Proprietary (API & app) | platform.claude.com |
| Claude Opus 4.7 | Apr 16, 2026 | 1M | 128K | $5 / $25 cached $0.5 New tokenizer from Claude 4.7: ~30% more tokens for the same text. | Text, image in → text out | — | Proprietary (API & app) | platform.claude.com |
| Claude Sonnet 4.6 | Feb 17, 2026 | 1M | 128K | $3 / $15 cached $0.3 | Text, image in → text out | — | Proprietary (API & app) | platform.claude.com |
| Claude Opus 4.6 | Feb 5, 2026 | 1M | 128K | $5 / $25 cached $0.5 | Text, image in → text out | — | Proprietary (API & app) | platform.claude.com |
| Claude Opus 4.5 | Nov 24, 2025 | 200K | 64K | $5 / $25 cached $0.5 | Text, image in → text out | May 2025 | Proprietary (API & app) | platform.claude.com |
Vendor benchmark claims
- Claude Opus 4.5: SWE-bench Verified 80.9% (vendor claim)
Benchmarks as published by the vendor at launch; not independently verified by i10x.
Sources: company announcements & release notes. Prices = standard API list price in USD per 1M tokens at time of release or current docs; batch is typically 50% off.
Try Claude in i10X
Compare it side by side with GPT, Gemini and Grok in one workspace.
Try Claude in i10X ›


