Anthropic gates Opus in Claude Code

// 45d agoPRODUCT UPDATE

Anthropic gates Opus in Claude Code

Anthropic’s Claude Code help center now states that Pro plan users can only use Opus models after enabling and purchasing extra usage. The same configuration guide still lets users switch models with `/model`, `--model`, or environment variables, but the access note makes the pricing boundary explicit: Sonnet and Haiku remain available, while Opus is reserved for users who opt into additional paid usage.

// ANALYSIS

This reads less like a model launch and more like a monetization guardrail around the highest-cost model tier.

–Pro users can still use Claude Code, but Opus access is no longer implied by the subscription alone.
–Anthropic is making the billing model explicit: if you want Opus in Claude Code, you need extra usage turned on and funded.
–The practical effect is to steer most day-to-day coding toward Sonnet/Haiku and reserve Opus for heavier or higher-value sessions.
–Power users may now compare the economics of Pro plus extra usage versus Max plans or API-key-based workflows.

// TAGS

anthropicclaude codeopuspro planextra usagepricingai coding

DISCOVERED

45d ago

2026-04-28

PUBLISHED

45d ago

2026-04-27

RELEVANCE

8/ 10

AUTHOR

Outside-Iron-8242

// KEEP READING

More AI developer news from the feed

EXPLORE FULL FEED

MODEL29m ago

Claude Mythos 5 Triggers White House Meeting

Anthropic's deployment of Claude Mythos 5 to vetted partners reportedly triggered a White House security meeting due to its advanced cybersecurity, biology, and chemistry capabilities. The publicly available version, Claude Fable 5, employs a routing safety mechanism to redirect sensitive queries and mitigate risks.

INFRA41m ago

Claude 5 runs on TPUs and Trainium

Tech commentator Theo Browne asked on X whether Anthropic's newly released Claude 5 models run on Google TPUs or AWS Trainium. In practice, Anthropic trains and deploys the models across Google Cloud TPUs, AWS Trainium, and NVIDIA GPUs as part of a diversified, cost-effective compute strategy.

MODEL59m ago

Anthropic launches public Claude Fable 5

Anthropic has publicly released Claude Fable 5, marking the general availability of its advanced Mythos-class model architecture for complex agentic workflows. While Fable 5 delivers frontier reasoning, it features active safety classifiers that route potentially sensitive prompts to Claude Opus 4.8 to ensure safety compliance.