OpenRouter Analysis Finds Opus 4.7 Tokenizer Increases Real World Costs

OpenRouter

Apr 28, 2026 · Updated May 6, 2026

OpenRouter's study of Opus 4.7 reveals that changes to the model's tokenizer have increased actual costs by 12% to 27% for most users. While short prompts have become more efficient, the shift highlights how token density can drive up expenses even when per-token pricing remains stable.

OpenRouter studied market data for the new Opus 4.7 and found that real-world costs increased by 12% to 27% for most users. This shift is driven by the model's tokenizer—the system that converts text into numerical tokens (the basic units of text processing). Short prompts were the only exception, showing improved efficiency.

This finding highlights a hidden variable in AI economics: token density. If a provider maintains the same price per million tokens, a less efficient tokenizer requires more tokens to represent the same sentence, effectively raising the price. This trend mirrors GitHub Copilot's usage-based billing as providers manage rising compute demands.

You should re-evaluate your API budget if your workflows rely on long-context prompts, as these now carry a significant cost increase. Conversely, applications using very short prompts may see slight cost improvements. These findings are based on OpenRouter's analysis and apply to their unified API.

View the full update on openrouter.ai

OpenRouter

@OpenRouterApr 28

We studied data across the market for Opus 4.7 and found that costs increased 12–27%, with the exception of short prompts, which actually got more cost efficient. Full post: https://t.co/c6Ypo9EglZ https://t.co/5D6v2z69LJ

1084

View on X

Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →

See all AI news & updates from OpenRouter →

Keep reading

OpenRouter Analysis Finds GPT 5.5 Conciseness Partially Offsets Double Pricing

OpenRouter analyzed real-world usage data and found that GPT-5.5's actual cost increase ranges from 49% to 92% despite OpenAI doubling per-token prices. While the model generates significantly fewer tokens for long-context tasks, users with shorter prompts face the full weight of the price hike without efficiency gains.

Simon Willison Finds Claude Opus 4.7 Tokenizer Raises Real API Costs

Simon WillisonApr 28

Simon Willison Finds Claude Opus 4.7 Tokenizer Raises Real API Costs

Simon Willison analyzed the new Claude Opus 4.7 tokenizer using his custom counter tool and found significant token inflation. While Anthropic kept per-token pricing the same, the increased token count for text and images creates a hidden price hike for API users.

Anthropic Increases Claude Subscriber Rate Limits to Offset Opus 4.7 Thinking

ClaudeApr 17

Anthropic Increases Claude Subscriber Rate Limits to Offset Opus 4.7 Thinking

Anthropic raised rate limits for all Claude subscribers to compensate for the higher token consumption of the new Opus 4.7 model. This adjustment ensures that the model's increased internal reasoning and updated tokenizer do not prematurely exhaust user quotas during complex tasks.

OpenCode adds Opus 4.7 to enable 1M context for open source agents

OpenCodeApr 16

OpenCode adds Opus 4.7 to enable 1M context for open source agents

OpenCode integrated the Opus 4.7 model to provide a 1-million-token context window for its open-source coding agent. This update allows developers to process massive codebases at the same price point as previous versions, making high-reasoning agentic workflows more accessible.