Skip to main content
Using Claude

Claude Model Comparison: Opus 5, Sonnet 5, Haiku 4.5, Fable 5.1

Compare specs and pricing for all four current Claude models (Fable 5.1, Opus 5, Sonnet 5, Haiku 4.5) and get clear selection criteria by task type, based on official docs.

ByUpdated
🌐 This article was machine-translated and may contain inaccuracies. Read the Korean original if in doubt.

The Claude API offers several models, and which one you choose drives both cost and quality. Based on the official Models overview (as of August 2026), this article compares the current models' specs and pricing and lays out selection criteria by task type.

🟢 Model references match the current lineup · model notice · Fable subscription
🟢 Model references match the current lineup · Claude Opus 5 / Claude Sonnet 5 / Claude Haiku 4.5 (higher tier: Claude Fable 5.1). This notice changes only when Anthropic ships a new model.

Fable 5 and 5.1 subscription (updated September 7, 2026): Claude Fable 5.1, released September 1, 2026, is the current Fable model and Fable 5 is now legacy. Plan terms are the same for both — Max and Team Premium plans include Fable at up to 50% of the weekly usage limit; Pro and Team Standard use usage credits (

🟢 Model references match the current lineup · Claude Opus 5 / Claude Sonnet 5 / Claude Haiku 4.5 (higher tier: Claude Fable 5.1). This notice changes only when Anthropic ships a new model.
0/M input, $50/M output tokens). The one-time
🟢 Model references match the current lineup · Claude Opus 5 / Claude Sonnet 5 / Claude Haiku 4.5 (higher tier: Claude Fable 5.1). This notice changes only when Anthropic ships a new model.
00 credit applied only to the Fable 5 transition and is not offered for 5.1. Some coding and debugging requests may be answered by an Opus model due to a security classifier (both models). See the Fable 5.1 guide and the Fable 5 availability guide for details.

Claude Model Comparison (September 2026) Pricing = USD per MTok · Official Models overview · Subject to change Fable 5.1 claude-fable-5-1 Price (in/out) $10 / $50 Context 1M Max output 128k Hardest frontier workloads Adaptive thinking (always on) Opus 5 claude-opus-5 Price (in/out) $5 / $25 Context 1M Max output 128k Complex reasoning & agentic coding Adaptive thinking supported Sonnet 5 claude-sonnet-5 Price (in/out)* $2 / $10 *Standard price (increase cancelled) Context 1M Max output 128k Best speed/intelligence balance Adaptive thinking (default on) Haiku 4.5 claude-haiku-4-5 Price (in/out) $1 / $5 Context 200k Max output 64k Fastest, cheapest bulk processing Extended thinking supported

Current model lineup

Per the official docs, the main current models are below. All current Claude models support text and image input, text output, multilingual, and vision, and are available via the Claude API, AWS, Bedrock, Vertex AI, and Microsoft Foundry.

  • Claude Opus 5 (claude-opus-5) — Anthropic's powerful Opus-tier model, for complex reasoning and long-horizon agentic coding. Pricing $5 input / $25 output per MTok, 1M context, 128k max output. Supports the effort parameter (defaults to high).
  • Claude Sonnet 5 (claude-sonnet-5) — the best balance of speed and intelligence, released June 30, 2026. The default model on Free and Pro plans, and available on Max, Team, and Enterprise. Pricing is $2 input / $10 output per MTok — originally announced as introductory pricing through August 31, 2026, but Anthropic cancelled the planned September 1, 2026 increase to $3 / $15 and made $2 / $10 the standard price (official pricing docs, checked September 2, 2026). 1M context, 128k max output. Adaptive thinking is on by default; extended thinking is not supported.
  • Claude Haiku 4.5 (claude-haiku-4-5-20251001, alias claude-haiku-4-5) — the fastest model with near-frontier intelligence. $1 / $5, 200k context, 64k max output. Supports extended thinking.
  • Claude Fable 5.1 (claude-fable-5-1) — Anthropic's most capable widely released model, for the most demanding reasoning and long-horizon agentic work. $10 / $50, 1M context, 128k max output. Adaptive thinking is always on. (Generally available from June 9, 2026.)

Note: Claude Mythos 5.1 (released September 1, 2026) follows the same model — Glasswing organizations only, with the same specs and pricing as Fable 5.1. Claude Mythos 5 / Mythos Preview are offered separately via Project Glasswing (invitation-only, defensive cybersecurity workloads) with no self-serve signup.

Pricing & spec comparison

Cost is priced per million tokens (MTok — one million units of text, roughly 750,000 words) of input/output. For the same task, unit price varies widely: Haiku $1/$5 → Sonnet 5 $2/$10 (intro) → Opus $5/$25 → Fable $10/$50. Output is pricier than input, so prompt design that reduces output length directly affects cost.

Update (September 2, 2026): Sonnet 5's $2/$10 was originally announced as introductory pricing through August 31, 2026, with a rise to $3/$15 from September 1. That increase did not happen — Anthropic's official pricing docs now state that $2/$10 is the standard price and the scheduled increase will not occur. Cost planning can use $2/$10 as the ongoing rate.

  • Context window — the maximum amount of text handled in a single request: Opus 5, Sonnet 5, Fable 5.1 = 1M tokens; Haiku 4.5 = 200k. (On Microsoft Foundry, Opus 4.8 is 200k.)
  • Max output: Opus 5, Sonnet 5 & Fable 5.1 = 128k; Haiku 4.5 = 64k. (Synchronous Messages API; on the Batch API some models support up to 300k output via a beta header.)
  • Thinking — the model's ability to reason internally before answering: Haiku 4.5 supports extended thinking (budget-specified). Opus 4.8, Sonnet 5, and Fable 5.1 use adaptive thinking (auto-adjusted). Sonnet 5 does not support extended thinking; Fable 5.1 has adaptive thinking permanently on.
  • Knowledge cutoff: Fable 5.1 Jun 2026, Opus 5 May 2026, Sonnet 5 Jan 2026, Haiku 4.5 Feb 2025 (official model docs, checked September 2, 2026).
  • New tokenizer: Sonnet 5 uses a different tokenizer from previous models. The same text produces roughly 30% more tokens than on Sonnet 4.6. If you migrated from Sonnet 4.6, recount your token budgets and cost estimates — do not reuse figures measured on older models.
Which model should I pick? Official guidance: when unsure, start with Opus 5 for complex tasks Task difficulty? Cost vs. quality trade-off Bulk · simple · low-latency → Haiku 4.5 Everyday production default → Sonnet 5 Complex reasoning · agents → Opus 5 Frontier-difficulty work → Fable 5.1 Key considerations · Long docs/codebases → 1M context (Opus · Sonnet 5 · Fable) · Extended thinking → Haiku 4.5 Adaptive → Opus · Sonnet 5 · Fable · Variable difficulty in a workflow → Route different models per step Always verify specs in official docs

Choosing by task type

The official docs advise "if unsure, start with Opus 5 for complex tasks." In practice, a cost/quality trade-off lens works well:

  • High-volume, simple, low-latency (classification, short replies, real-time chatbots) → Haiku 4.5. Fastest and cheapest.
  • Everyday production default (most coding, generation, summarization) → Sonnet 5. Balanced speed, intelligence, and cost. It is the default model on Free and Pro plans.
  • Complex reasoning & long-horizon agentic coding → Opus 5. When answer quality matters most.
  • Frontier-difficulty work → Fable 5.1. Only when the extra cost is justified by the difficulty of the task.

For long documents or large codebases, the 1M context window (Opus, Sonnet 5, Fable) is an advantage. For workflows where difficulty shifts step to step, routing different models per step is a practical approach.

Model IDs are pinned snapshots

A caveat: every Claude model ID is a pinned snapshot — think of it like a frozen recipe card, not a live feed. IDs with a date (e.g., 20251001) are fixed to that release, and starting with the 4.6 generation the dateless format (e.g., claude-sonnet-5) is also a pinned snapshot, not an evergreen pointer that silently updates. You can query model capabilities and token limits programmatically via the Models API (returns max_input_tokens, max_tokens, and a capabilities object).

Summary

A sensible model-selection flow: (1) gauge task difficulty and cost tolerance → (2) pick a tier among Haiku / Sonnet 5 / Opus / Fable → (3) fine-tune by context window, thinking mode, and output needs. To learn more, see Claude API Model IDs and Versioning and Context Window and Tokens.

This article summarizes public information from the official Anthropic docs (platform.claude.com/docs) as of August 2026. The model lineup, pricing, and specs change frequently, so always verify against the official Models overview and Pricing pages when choosing. This site is not an official Anthropic site.

Was this helpful?

Keep reading

Readers read this next

Opus vs Sonnet: Which Claude Model Should You Use?

Claude Opus 5 and Sonnet 5 differ in price, speed, context window, and thinking mode. Here's how to pick the right one in the API or the Claude.ai app.

301

Claude API Pricing by Model — Input/Output Token Rates and How to Cut Costs with Batch and Prompt Caching

Compare Claude API token pricing across Opus 5, Sonnet 5, Haiku 4.5, and Fable 5.1, and learn how Batch API's 50% discount and prompt caching's up to 90% discount can lower your costs.

86

Can You Use Claude to Search Korean Law and Court Rulings? What It Can and Can't Do

Claude has no direct connection to Korea's official law and court case databases and can invent case numbers that don't exist. Here's the safer way to use it — pasting in the real text yourself.

1

Anthropic's Claude Misuse Reports: The Major Cases Explained

Anthropic's threat intelligence reports document real cases of Claude being misused for hacking, fraud, and influence operations — and how Anthropic detected and stopped them.

0