Claude: pricing, limits, errors and comparisons
Anthropic's Claude models and apps: the Claude API, model choices, subscription plans and usage limits. Claude Code has its own page.
32 articles · last updated Oct 8, 2026
Pricing and costs
- Claude API Pricing Per Million Tokens (October 2026): $0.10–$50
Haiku 5.5 is the cheapest Claude model until a prompt passes 100K tokens; then every rate is 5x. Batch halves any price; cache hits cost 2.5–10% of input.
Updated
- Claude Cowork Pricing: What You Actually Pay on Each Plan
There's no Cowork add-on to buy. Pick the plan whose allowance fits how much work you hand off, and know what extra usage costs before you hit the limit.
Updated
- Claude Opus 4.6 Pricing: Complete Guide to API Costs, Subscriptions & Savings (2026)
Complete pricing guide for Claude Opus 4.6 — Anthropic's flagship model at $5/$25 per million tokens. Covers standard, fast mode ($30/$150), 1M context beta, Batch API (50% off), prompt caching (90% savings), subscription plans, and competitive comparison vs GPT-5.2 and Gemini 3 Pro.
Updated
Free tiers and limits
- Claude API Quota Tiers and Limits Explained: Start, Build, Scale
Anthropic assigns usage tiers automatically: Start caps monthly spend at $500, Build at $1,000, Scale at $200,000. Cache reads don't count toward ITPM.
Updated
- Does Claude Pro Have a Free Trial? How to Check Your Account in 2026
There is no Claude Pro trial you can start on demand. Free is a standing $0 plan. Pro at no cost comes only by promotion, gift, or a shared guest pass.
Updated
- Claude Fable 5.1 Usage Limits: Weekly Allowance, Extra Usage, and Model Switching
Max includes Fable within its shared weekly allowance; Pro requires credits from the first request. Check the relevant limit and your credit settings before continuing.
Updated
- Claude Opus 5.5 Limit Reset: When to Spend It
Read which limit your card restores, spend a weekly card when your regular reset is days away, and keep heavy work on Opus 5.5 unless it needs Fable 5.1.
Updated
Errors and fixes
- How to Fix Claude API 429 Rate Limit Error: 8 Causes and Fixes
Check retry-after first. If present, wait, add jitter and pace traffic. If absent with enforced_spend_limit_reached, stop until a tier increase or the 1st.
Updated
- Claude API Error 500: Fix api_error Without Blind Retries
Recover from a returned Claude internal server error by checking status, saving request IDs, retrying with a bounded budget, verifying the same path, and escalating with evidence.
Updated
- Claude 529 Overloaded Error: What It Means and How to Fix It (2026)
Claude 529 overloaded_error usually means Anthropic is temporarily overloaded or an upstream service is saturated. This guide explains how to verify that quickly, how it differs from 429 and Claude chat capacity warnings, and how to reduce repeat failures.
Updated
Comparisons and alternatives
- Claude Haiku 5.5 vs GPT-6 Luna Pricing: Same Until 100K, Then 5x
Identical bills up to 100,000-token prompts; above that, Luna is 5x cheaper per request until 272K. Below the line, compare cost per task at matched quality.
Updated
- Claude Sonnet vs Opus vs Haiku vs Fable: Which Model to Use
Haiku 5.5 is the fastest, cheapest Claude model ($0.10/$0.50 up to 100K-token prompts) and Fable 5.1 the most capable. Where to start depends on how you pay.
Updated
- Gemini API vs OpenAI vs Claude: Complete 2026 Cost Decision Guide
Gemini is no longer cheapest at every tier: Flash-Lite, GPT-6 Luna and Haiku 5.5 start at $0.10 input, and Sol and Sonnet 5.5 undercut Gemini 3.1 Pro on output.
Updated
- GPT-6.1 Sol vs Claude Opus 5.5: Half the Price, 1/6 the Task Cost
Try GPT-6.1 Sol first for high-volume work under 272K input tokens. Keep Opus 5.5 where a failed attempt costs more review time than the tokens you save.
Updated
- Gemini 4 Argon vs Claude Opus 5.5 vs Sonnet 5.5: Which to Use
As of October 1, 2026, only the two Claude models can be called. Use Sonnet 5.5 for well-scoped work, Opus 5.5 for long open-ended jobs, and re-test Argon later.
Updated
- Claude Opus 5.5 vs GPT-5.6 Sol: Same Price Tag, Different Bills
Opus 5.5 costs less above 272K input tokens and on cache reads; GPT-5.6 Sol writes about half the output per task. Already on Sol? Test GPT-6 Sol or the newer GPT-6.1 Sol first.
Updated
- GPT-6 Astra vs Claude Fable 5.1 API Cost: Where the Tie Breaks
Their standard input and output rates match. Cache behavior, Astra's 272K threshold, batch processing, tools, and your token mix determine the actual bill.
Updated
- GPT-5.4 vs Claude Sonnet 4.6: Which Should You Use in 2026?
GPT-5.4 is the better default for most developers who want one model for coding, long-context work, and tool-heavy agents in March 2026. Claude Sonnet 4.6 still has a real place in Claude-first workflows, but its base published API pricing is not cheaper than GPT-5.4.
Updated
- Claude Opus 4.6 vs Sonnet 4.6: Which Should You Use in 2026?
Claude Sonnet 4.6 is the better default for most users in March 2026, while Claude Opus 4.6 is the justified upgrade for harder reasoning, longer autonomous work, and larger single-pass outputs. This guide compares pricing, context, access, and workflow fit using current Anthropic sources.
Updated
- Gemini 3.1 Pro vs Opus 4.6 vs Codex 5.3: The Definitive Comparison (March 2026)
Three frontier models released in February 2026 have fundamentally changed the AI landscape. This guide compares Gemini 3.1 Pro, Claude Opus 4.6, and GPT-5.3-Codex with browser-verified pricing, real benchmark data, and practical recommendations for developers choosing their next production model.
Updated
- GLM-5 vs Claude Opus 4.6 vs GPT-5.3 Codex: The Definitive Three-Way Comparison (2026)
As of February 2026, three heavyweight AI models launched in the same week: Claude Opus 4.6 ($5/MTok, best reasoning), GPT-5.3 Codex (~$1.75/MTok, fastest terminal tasks), and GLM-5 (~$0.11/MTok, open-source 745B MoE). This guide provides the industry's first complete three-way comparison with verified benchmarks, real-world cost analysis, and a clear decision framework.
Updated
Guides and setup
- Claude Startup Credits: Who Qualifies and What $1,000 Buys
Eligible if founded in the last 5 years or funded in the last 2. $1,000 covers ~71,000 Sonnet 5.5 calls at 3,000 in/800 out tokens; Team renews at full price.
Updated
- Claude Opus 4.6 API Key: Setup, Integration and Production Deployment
Complete one-stop guide to obtaining and integrating your Claude Opus 4.6 API key. Covers account setup, Python and Node.js SDK integration, security best practices, rate limit handling, and cost optimization strategies. Includes production-ready code examples verified February 2026.
Updated
- How to Use Prompt Caching in Claude API: Complete 2026 Guide with Code Examples
Master Claude API prompt caching with this comprehensive guide. Get production-ready code examples in Python and TypeScript, detailed cost calculations, and expert troubleshooting tips to optimize your API usage.
Updated
- Claude Image Generation Connector: Which to Add, What It Costs
A connector hands your Claude prompt to an outside image service. Choose by who runs it, the account it needs, and its price. Free plans get one custom connector.
Updated
- Claude Mythos Access: Who Can Get Mythos 5.1 and How to Apply
Mythos Preview is now Mythos 5.1, still with no signup, waitlist, or free tier. Only approved organizations get it; Fable 5.1 is the same model with safeguards.
Updated
- Claude Opus 5.5 Effort: Choose a Level, Set It, and Measure the Cost
Start at medium for Opus 5.5, then move effort only when your own completed tasks justify the extra thinking, time, and cache cost.
Updated
- Claude Opus 5.5: API Price, Model ID, and Migration Checks
Opus 5.5 has a lower direct API token rate than Opus 5, but an existing Messages API client may need more than a model ID change. Here is the price math and the compatibility check.
Updated
- Claude Fable 5.1: API Pricing, Changes, and Migration Guide
Fable 5.1 does not lower Fable's base token rates. It cuts cache-read cost by 75% and introduces tool-choice and thinking-history rules that can break an existing client.
Updated
- Move ChatGPT or Gemini Memory to Claude—and Use Takeout Correctly
For Gemini-to-Claude moves, use Claude's prompt to create a reviewed memory handoff. Use Takeout for history and backup—not as a raw memory import.
Updated
- How to Get a Claude API Key Without Buying Someone Else’s Secret
Do not pay for a loose key string. Pay for access in an account you control, confirm which provider bills the request, and prove the route with a small auditable call.
Updated
- Is There a Claude Student Discount in 2026? The Routes That Actually Exist
A student email does not automatically discount Claude Pro. The useful options are public individual plans, access sponsored by a participating university, and separate student applications.
Updated