Anthropic attributes Claude Code usage drain to peak-hour caps and large context windows
Anthropic has identified two primary causes for Claude Code users hitting usage limits faster than expected: stricter rate limiting during peak hours and sessions with context windows exceeding 1 million tokens. The company also recommends switching to Sonnet 4.6 instead of Opus, which consumes limits roughly twice as fast.
Anthropic Attributes Claude Code Usage Drain to Peak-Hour Caps and Expanding Contexts
Anthropichas identified the root causes behind complaints from Claude Code users depleting their usage limits faster than anticipated, according to Anthropic's Lydia Hallie.
The two primary factors are:
- Tighter peak-hour rate limits — Anthropic implements stricter usage caps during periods of high demand
- Ballooning context windows — Sessions with 1-million-token contexts or larger consume limits significantly faster
Bug Fixes and Billing Accuracy
Hallie confirmed Anthropic fixed several bugs but stated none resulted in incorrect billing. The company has deployed efficiency improvements and added in-product notifications to help users understand their usage patterns better.
Anthropic's Recommendations
To manage usage more effectively, Hallie recommends:
- Switch to Sonnet 4.6 instead of Opus, which consumes limits approximately twice as fast
- Disable Extended Thinking when not required for specific tasks
- Start fresh sessions rather than continuing previous conversations to reduce accumulated context
- Limit context window size to reduce per-request token consumption
Users experiencing usage depletion they believe is unusual should report the issue through the in-product feedback function.
What This Means
The clarification addresses a widespread concern among Claude Code users and reveals that usage consumption is largely a function of rate limiting mechanics and user behavior rather than billing errors. The recommendation to use Sonnet 4.6 over Opus suggests a significant performance-per-token trade-off between the two models. Anthropic's focus on user transparency through pop-ups and clearer guidance indicates the company is attempting to manage expectations around usage consumption before users encounter hard limits.
Related Articles
Google opens CodeMender API to select testers, pitching AI security tool to governments and enterprises
Google announced at I/O 2026 that it is opening API access for CodeMender, its AI agent for code security, to select expert groups. The company is positioning the tool to compete with Anthropic's Mythos Preview, which flagged unknown security vulnerabilities and secured major government and enterprise contracts.
AWS releases four multimodal evaluators for image-to-text AI tasks in Strands Evals SDK
AWS has added four multimodal evaluators to its Strands Evals SDK that judge image-to-text AI outputs by directly analyzing source images. The evaluators—Overall Quality, Correctness, Faithfulness, and Instruction Following—use multimodal large language models to detect visual hallucinations, factual errors, and instruction violations that text-only judges miss.
AWS SageMaker AI adds bidirectional streaming for real-time speech transcription with vLLM
Amazon SageMaker AI has launched bidirectional streaming support for real-time inference, enabling WebSocket-based voice applications through vLLM integration. The feature uses HTTP/2 on port 8443 to bridge client connections with vLLM's Realtime API, allowing audio to stream in while transcription streams back simultaneously over a single persistent connection.
Google launches Universal Cart, an AI agent that shops across multiple retailers in one checkout
Google announced Universal Cart at its I/O developer conference, an AI-powered shopping system that consolidates purchases from multiple retailers including Target, Shopify, Wayfair, and Etsy into a single checkout. The feature uses Gemini's agentic AI to verify product compatibility, suggest better deals, and automate routine purchases.
Comments
Loading...