November 24, 2025
Cheap genius or nerfed rival?
Claude Opus 4.5
Cheaper Claude Opus 4.5 has devs cheering—and accusing a Sonnet "nerf"
TLDR: Anthropic launched Claude Opus 4.5 with a hefty price cut and top coding scores. Devs are thrilled about cheaper power but split over claims Sonnet got nerfed; they’re asking for Claude Code access and higher limits, because affordable, capable AI could reshape how they code and work.
Anthropic just dropped Claude Opus 4.5 and the crowd went full popcorn mode. The company says it’s their smartest coding and “agent” model yet, now priced at $5/$25 per million tokens—think of tokens as the AI’s word budget you pay for. Commenters like jumploops are hyped: “1/3 the price of Opus 4.1… and uses a lot less tokens?” Meanwhile elvin_d is doing wallet math, noting it’s close to Google’s Gemini Pro pricing and begging for higher usage limits.
Bench charts show Opus 4.5 topping SWE-bench Verified—a tough real-world coding test—so naturally the “but can it ship?” crowd is celebrating. bnchrch sighs in relief, saying these scores mean they don’t have to rip out their coding setup: FOMO is exhausting, they confess, and they’re keeping Claude Code like a favorite coffee mug. The vibe: excitement meets burnout.
Then the drama hits. stavros claims Sonnet 4.5 feels “much dumber” lately and wonders if that’s timed to make Opus look better. The thread splits between “cheap, smart, go!” and “did they nerf the old one?” Also: people keep asking if Opus 4.5 works in Claude Code—Anthropic says it’s available across apps and the API, with fresh updates to Claude Code, Excel, Chrome, and desktop. Expect stress tests, wallet memes, and lots of “my coding agent lives” jokes.
Key Points
- •Anthropic released Claude Opus 4.5, positioning it as its most capable model for coding, agents, and general computer use.
- •Opus 4.5 is available via Anthropic apps, the Claude API (model ID 'claude-opus-4-5-20251101'), and major cloud platforms.
- •Pricing is set at $5 per million input tokens and $25 per million output tokens.
- •Anthropic highlights state-of-the-art performance on SWE-bench Verified and improved handling of complex tasks versus Sonnet 4.5.
- •Updates accompany the release: enhanced Claude Developer Platform, Claude Code, and apps with tools for longer-running agents and integrations for Excel and Chrome; longer conversations are supported.