On the heels of Microsoft Construct and Google I/O, Anthropic has simply introduced Claude 4 Sonnet and Claude 4 Opus, that are instantly accessible on Claude’s web site, in addition to within the API. Right here’s what’s new.
Higher at coding and at … Pokémon
In response to Anthropic, Claude Sonnet 4 (its mid-tier mannequin, between Raiku and Opus) considerably improves at coding, reasoning, and instruction following in comparison with its predecessor, Claude Sonnet 3.7.
As for Claude Opus 4, Anthropic says it matches or outperforms OpenAI’s o3, GPT-4.1, and Gemini 2.5 Professional in benchmarks for multilingual Q&A, agentic software use, agentic terminal coding, agentic coding, and graduate-level reasoning:
That is particularly vital as a result of, whereas Claude spent most of final yr on the prime of builders’ most popular fashions for coding duties, it has fallen behind in latest weeks after a number of mannequin updates by OpenAI and Google.
And talking of Google, its Gemini 2.5 Professional mannequin made the rounds not too long ago after it accomplished a Pokémon Blue playthrough. Anthropic was joyful to report that whereas it hasn’t but achieved the identical feat, Claude Opus 4 was capable of agentically play Pokémon for 24 hours, versus 45 minutes from the earlier model.
New capabilities and Claude Code
Alongside the fashions, Anthropic additionally introduced:
• Prolonged considering with software use (beta): Each fashions can use instruments—like internet search—throughout prolonged considering, permitting Claude to alternate between reasoning and power use to enhance responses.
• New mannequin capabilities: Each fashions can use instruments in parallel, observe directions extra exactly, and—when given entry to native recordsdata by builders—exhibit considerably improved reminiscence capabilities, extracting and saving key information to keep up continuity and construct tacit data over time.
• Claude Code is now typically accessible: After receiving intensive constructive suggestions throughout our analysis preview, we’re increasing how builders can collaborate with Claude. Claude Code now helps background duties through GitHub Actions and native integrations with VS Code and JetBrains, displaying edits straight in your recordsdata for seamless pair programming.
• New API capabilities: We’re releasing 4 new capabilities on the Anthropic API that allow builders to construct extra highly effective AI brokers: the code execution software, MCP connector, Information API, and the power to cache prompts for as much as one hour.
The Claude Code information is especially attention-grabbing for builders, since @ mentioning Claude and letting it run straight from a GitHub PR has the potential to streamline the event course of.
Anthropic says each fashions can be found on the Anthropic API and companions like Amazon Bedrock and Google Cloud’s Vertex AI. Opus 4 prices $15/$75 per million tokens (enter/output), and Sonnet 4 prices $3/$15 per million tokens (enter/output).
Do you employ Claude or different LLMs at work? Tell us within the feedback.
FTC: We use earnings incomes auto affiliate hyperlinks. Extra.
