Skip to content
gentic.news — AI News Intelligence Platform
Connecting to the Living Graph…

token budget

30 articles about token budget in AI news

Scaling Law Plateau Not Universal: More Tokens Boost Reasoning AI Performance

Empirical evidence indicates the 'second scaling law'—performance gains from increased computation—does not fully plateau for many reasoning tasks. Benchmark results may be artificially limited by token budgets, not model capability.

85% relevant

SemiAnalysis: Perplexity Slack Bot Beats Claude in Internal Trial

SemiAnalysis found Perplexity's Slack bot beats Claude in internal trial. 96% token budget goes to Anthropic, but usage may shift.

75% relevant

NVIDIA Spending ~$75K Per Engineer on AI Compute Tokens, Indicating Multi-Billion Dollar Annual Budget

NVIDIA is reportedly allocating approximately $75,000 in AI compute tokens per engineer annually, translating to a multi-billion dollar organization-wide budget for AI development resources.

87% relevant

AISI: Fixed compute budgets underestimate AI agents by 60%

AISI found standard benchmarks cap compute budgets, underestimating agent capabilities by ~60%. Success rates jumped ~25% with 10x tokens.

96% relevant

How to Prevent Cost Explosions with MCP Gateway Budget Enforcement

Standard MCP gateways miss economic governance. Add per-tool cost modeling and budget-aware tokens to prevent agents from burning through thousands in minutes.

85% relevant

LongStraw Reaches 2.1M Tokens on 8 H20 GPUs via Branch Replay

LongStraw reaches 2.1M token positions for RL post-training on 8 H20 GPUs by replaying short response branches, cutting compute 8-16x vs prior art.

87% relevant

NVIDIA's #1 RTEB Embedding Model Skips Token Generation Entirely

NVIDIA's RTEB hit #1 on MTEB by skipping token generation. The cheapest reasoning token is the unused one.

83% relevant

UniSound U2 Cuts Token Use 25%, Joins Top Chinese LLM Tier

UniSound's U2 foundation model cuts token consumption by 25% while matching top Chinese LLM performance, entering the top tier with an efficiency-first design.

71% relevant

Claude Code Token Costs Got You Down? Here's How to Cut Usage 40% Without

Claude Code users frustrated by token costs should use /compact, optimize CLAUDE.md, and route cheap models via OpenRouter for simple tasks—no local model matches Claude's quality yet.

90% relevant

Anthropic Deprecates Fixed Thinking Budgets, Forces Adaptive Mode

Anthropic forced adaptive thinking on Claude models, deprecating fixed budgets. Users report quality drops and the change reduces API revenue potential.

100% relevant

Doby Cuts Claude Code Navigation Tokens by 95% with Spec-First Workflow

A spec-first fix workflow that slashes navigation tokens 95% and enforces plan docs as source of truth before code changes.

100% relevant

Meta Employee Builds 'Claudeonomics' Dashboard for Internal AI Token Competition

A Meta employee built an internal dashboard called 'Claudeonomics' that ranks coworkers by their usage of company AI tokens, creating a gamified competition and providing a novel view into internal AI tool adoption patterns.

75% relevant

Uber's AI Budget Blowout Is a Warning for Every Claude Code User

Uber's experience shows unmanaged Claude Code usage can explode costs. Developers must implement usage tracking and set clear per-task budgets.

100% relevant

Codeburn: The TUI That Shows Exactly Where Your Claude Code Tokens Are Going

A new open-source TUI, Codeburn, analyzes Claude Code session transcripts to show token spend by task type, helping developers optimize their usage and costs.

100% relevant

Stop Wasting Your CLAUDE.md Instruction Budget — Here's What Actually Works

Your CLAUDE.md has a limited instruction budget. Stop filling it with platitudes and obvious stack info. Focus on project-specific failure modes and safety guards.

95% relevant

AI Efficiency Breakthrough: New Framework Optimizes Agentic RAG Systems Under Budget Constraints

Researchers have developed a systematic framework for optimizing agentic RAG systems under budget constraints. Their study reveals that hybrid retrieval strategies and limited search iterations deliver maximum accuracy with minimal costs, providing practical guidance for real-world AI deployment.

79% relevant

Claude Opus 4.7 Launches with 3.75MP Vision, Agentic Coding, and New Tokenizer

Anthropic launched Claude Opus 4.7 today with 3x higher vision resolution (3.75MP), self-verifying coding outputs, and stricter instruction following. The update targets enterprise agentic workflows and knowledge work benchmarks.

100% relevant

Claude Sonnet 4.6 Is Live: How to Use the New 'Budget Flagship' Model in Claude Code

Anthropic's new Claude Sonnet 4.6 model offers near-Opus performance at a Sonnet price. Here's how to configure Claude Code to use it for maximum efficiency.

89% relevant

How Claude Code's New API Pricing Changes Your Development Budget

Anthropic's new API pricing tiers mean you can now use Claude Code for more tasks without breaking the bank. Here's how to adjust your usage.

95% relevant

Claude Code's New /review Command: How to Use It Without Breaking Your Budget or Team

Claude Code now has built-in code review. Learn the exact prompts and CLI flags to make it cost-effective and complementary to senior engineers.

96% relevant

Claude Fable 5 Migration: Cut Prescriptive Skills 60% to Stop Degrading Output

Audit your ~/.claude/skills for temperature, budget_tokens, and 'show your reasoning'. Replace 6+ step procedures with goal+constraints. Cut MUST/NEVER blocks to only guard money, deletions, or identity.

100% relevant

Agent Harness Scaling: EFC Predicts Success at R2 0.99 vs 0.42

New research introduces Effective Feedback Compute (EFC), which predicts agent success at R2 0.99 vs 0.42 for raw tokens. Reallocating compute by EFC lifts success 3x at the same budget.

88% relevant

Claude Code Max Users Report 3–4x Faster Limit Exhaustion

Max plan users see 3–4x faster limit exhaustion. Audit token-heavy tasks, use /compact and CLAUDE.md rules to reduce waste per session.

90% relevant

LLMs Learn to Switch Reasoning Effort at Inference Time

@rasbt explains how LLMs switch reasoning effort using inference-time methods and training, potentially cutting token usage by 30–50% on simple queries.

75% relevant

The One Constraint That Makes Claude Code Prompts Work (Or Fail)

Protect Claude Code's context window budget: be specific, provide a verifiable check (tests, build), use plan mode for multi-file changes, and keep CLAUDE.md lean. This one constraint drives all prompt best practices.

90% relevant

DeepSeek, Zhipu AI Build Custom Inference Chips to Cut GPU Dependency

DeepSeek and Zhipu AI are developing custom inference chips to cut GPU costs. China's domestic chip budget share hit 46% in July 2026.

100% relevant

Claude Code Digest — Jul 04–Jul 07

Agentic coding is getting more expensive to debug than to generate: Lovable burned $85K in tokens, and that’s the part enterprises keep underestimating.

95% relevant

AI Security Inst Shows Test-Time Compute Skews Frontier Evaluations

AISecInst research shows test-time compute budgets skew frontier model evaluations, challenging standard practices.

92% relevant

Claude Fable 5 in Claude Code: The Routing Strategy That Saves Your Weekly Limit

Claude Fable 5 ($10/$50 per M tokens) scores 91 on Senior Engineer benchmarks vs Opus 4.8's 63—use `/model fable` for complex, multi-file tasks, but reserve quick edits for cheaper models to save your weekly limit.

88% relevant

Cursor's Composer 2.5 matches Opus 4.7, GPT-5.5 at fraction of cost

Cursor's Composer 2.5 scores 79.8% on SWE-Bench Multilingual at $0.50/M tokens, matching Opus 4.7 and GPT-5.5 at 30x lower cost.

95% relevant