Introducing Claude Opus 4.6. Our smartest model got an upgrade. Opus 4.6 plans more carefully, sustains agentic tasks for longer, operates reliably in massive codebases, and catches its own mistakes. It’s also our first Opus-class model with 1M token context in beta.

Feb 5, 2026 · 5:45 PM UTC

1,658
4,644
38,736
10,621,941
Opus 4.6 is state-of-the-art on several evaluations including agentic coding, multi-discipline reasoning, knowledge work, and agentic search. We're also shipping new features across Claude in Excel, Claude in PowerPoint, Claude Code, and our API to let Opus 4.6 do even more.
Benchmark comparison table showing Claude Opus 4.6 leading or matching top scores across 15 evaluations — including agentic coding, tool use, reasoning, and search — against Opus 4.5, Sonnet 4.5, Gemini 3 Pro, and GPT-5.2.

ALT Benchmark comparison table showing Claude Opus 4.6 leading or matching top scores across 15 evaluations — including agentic coding, tool use, reasoning, and search — against Opus 4.5, Sonnet 4.5, Gemini 3 Pro, and GPT-5.2.

97
303
4,062
751,759
Claude in Excel now handles long-running and harder tasks with improved performance. It can plan before acting, support richer functionalities like conditional formatting and data validation, and handle multi-step changes in one pass. Read more: claude.com/blog/opus-4-6-fin…
21
129
2,489
758,088
Claude in PowerPoint is now available in research preview for Max, Team, and Enterprise. Claude reads your layouts, fonts, and slide masters to stay on-brand — whether you're building from a template or generating a full deck from a description.
Claude's PowerPoint integration shown inside PowerPoint, with an Opus 4.6 chat panel analyzing a selected slide's EV charging market chart and providing key data points alongside suggested edits.

ALT Claude's PowerPoint integration shown inside PowerPoint, with an Opus 4.6 chat panel analyzing a selected slide's EV charging market chart and providing key data points alongside suggested edits.

29
156
2,647
941,495
On Claude Code, we’re introducing agent teams. Spin up multiple agents that coordinate autonomously and work in parallel—best for tasks that can be split up and tackled independently. Agent teams are in research preview: code.claude.com/docs/en/agen…

Orchestrate teams of Claude Code sessions - Claude Code Docs

Coordinate multiple Claude Code instances working together as a team, with shared tasks, inter-agent messaging, and centralized management.

code.claude.com
87
348
3,991
865,049
New on the API: we're giving developers better control over model effort and more flexibility for long-running agents. Adaptive thinking lets Claude calibrate its reasoning depth to each task, and context compaction keeps long-running tasks from hitting limits.
20
34
1,082
268,989
Claude Opus 4.6 is available today on claude.ai, the Claude Developer Platform, and all major cloud platforms. And within Cowork, Opus 4.6 can put all these skills to work autonomously on your behalf. Read more: anthropic.com/news/claude-op…

Claude

claude.ai
56
92
1,617
260,328
Sort replies: Relevant Recent Liked
Replying to @claudeai
You are here
23
43
1,183
64,006
Replying to @claudeai
4
9
385
27,398
Replying to @claudeai
hmm..
11
1
343
61,120
Replying to @claudeai
Opus 4.6 doesn't like being a product
17
13
325
23,634
Replying to @claudeai
Ok so we're trusting the AI-that-we-admit-could-be-ASL 4-and-kill-everyone to grade itself on safety, because humans can't keep up anymore This is fine 👍
17
3
214
20,079
Replying to @claudeai @AnthropicAI
Best coding model I’ve ever used We will have a detailed vibe check + livestream on @every in 15 minutes: every.to

Every

Every — The only subscription you need to stay at the edge of AI. Ideas, apps, and training from practitioners who build with AI daily.

every.to
11
7
199
44,229
Replying to @claudeai
sup guys, got this new lamborghini here
1
5
199
11,207
Replying to @claudeai
Finally 1M context on Claude models
5
8
332
37,601
Replying to @claudeai
are you guys fooling us
5
162
45,013
Replying to @claudeai
?? Apollo just... didn't actually safety test Opus 4.6? It knows it's being tested, so why bother?
5
8
132
16,512
Replying to @claudeai
2
135
7,020
Replying to @claudeai
Very interesting that it's .1% lower on SWE-Bench yet MUCH higher on many other benchmarks. I wonder what the tradeoff is there?
6
107
18,208
Replying to @claudeai
1M context let's goo!
5
80
8,876
Replying to @claudeai
i don’t care about Opus this is the best thing Anthropic ever built
Claude
@claudeai
Feb 4
Ads are coming to AI. But not to Claude. Keep thinking.
1
77
10,125
Replying to @claudeai
Feels like Claude decided to mog sama's announcement:
72
6,039
Replying to @claudeai
Opus Testing Time! 🔥
1
68
4,741
Replying to @claudeai
Holy shit
4
67
3,351
Replying to @claudeai
1
58
6,079
Replying to @claudeai @AnthropicAI
AMAZING! This will make Claude Code way more powerful. Here’s the easiest way to get started - it’s a complete course IN Claude Code so you learn by DOING, not reading or watching ccforeveryone.com 100% free (includes vibe coding + cowork)

Claude Code Tutorial (Free Course) | No Coding Required – CC for Everyone

Learn Claude Code IN Claude Code! Free Claude Code tutorial — no coding required. Master AI agents, file ops, and vibe coding. Hands-on, no videos.

ccforeveryone.com
1
5
38
12,110
Replying to @claudeai
when will it end
35
2,678
Replying to @claudeai
Claude Opus 4.6 is SOTA on the Vals Index, and also tops out our Finance Agent, ProofBench, Swe Bench, Tax Eval leaderboards.
Vals AI
@ValsAI
Feb 5
Claude Opus 4.6 is #1 on the Vals Index 🏆 It sets a new state-of-the-art on FinanceAgent, ProofBench, TaxEval, and SWE-Bench. (1/n)
35
5,006
Replying to @claudeai
lovin' it folks awesome work!
Zaid
@zaidmukaddam
Feb 5
Claude Opus 4.6 is now available on scira.ai!
1
32
3,521
Replying to @claudeai
1M context on opus-class is the real news here. "catches its own mistakes" better mean it actually stops and re-reads instead of confidently bulldozing forward like current opus does
5
32
8,884
Replying to @claudeai
FINALLY THANK U SO MUCH CLAUDE
21
1,316
Replying to @claudeai
Live Vibe Check on all things Opus 4.6:
Dan Shipper
@danshipper
Feb 5
Vibe Check Live: GPT-5.3 Codex vs. Opus 4.6—INSANE new models from @OpenAI and @AnthropicAI. We’ve been testing both hands-on for over a week. Let's go through our head-to-head take! x.com/i/broadcasts/1vAxRQOyr…
2
1
20
6,336
Replying to @claudeai
23
580
Replying to @claudeai
Finally the context window is solved Was super annoying to claude asking to open new conversation due to context windows And the chat compression was terrible!
21
1,808
Replying to @claudeai
Let's GO! Covering this live here: x.com/i/broadcasts/1kvJpMbEw…

Alex Volkov

📆 Feb 5 - Clankers internet, Codex has an App, AceStep suno at home, Kling vs Grok & more AI news
1
15
7,219
Replying to @claudeai
opus with 1M token context window? our @openclaw got an upgrade...
1
16
6,237
Replying to @claudeai
This is about to get wild! Sonnet 5 next??
16
4,482
Load more