Anthropic now sells four tiers at once — Haiku 4.5, Sonnet 5, Opus 4.8 and Fable 5 — spanning 10× on input price and 50× end-to-end. Treating that menu as a single choice is how teams overpay. Treating it as a scheduling problem is how platforms win. Verified prices, honest positioning, a decision flow, and a live cost simulator below.
Prices from Anthropic’s official pricing page, retrieved July 4, 2026. Sonnet 5 introductory pricing runs through August 31, 2026. Notice the rhyme: Opus 4.8’s input costs exactly what Haiku 4.5’s output costs — the ladder is deliberately spaced.
Anthropic no longer ships a single flagship. It ships a price-segmented ladder, and every tier has a legitimate job. The one most comparisons skip — Haiku — is the one that should be carrying most of your call volume.
The fastest Claude, with near-frontier intelligence. The only current model on the previous tokenizer and a 200k window — and priced for anything you run millions of times a day.
The best balance of speed and intelligence in the lineup — the most agentic Sonnet yet, built to plan and drive tools autonomously. Launched June 30, 2026.
Built for complex agentic coding and enterprise work. Effort defaults to high on every surface, and a research-preview fast mode roughly doubles the rate when latency matters. Launched May 28, 2026.
Next-generation intelligence for long-running agents — the generally available half of the Mythos-class tier, shipped with dual-use safety classifiers (Mythos 5 is the invitation-only variant). Adaptive thinking is always on. GA June 9, 2026.
Everything below is from Anthropic’s official model and pricing documentation. The rows most people skip — cache-read rates and the tokenizer generation — are the ones that decide real-world cost.
| Spec | Haiku 4.5 | Sonnet 5 | Opus 4.8 | Fable 5 |
|---|---|---|---|---|
| Positioning | Fastest, near-frontier intelligence | Best speed-to-intelligence balance | Complex agentic coding & enterprise work | Next-gen intelligence for long-running agents |
| Input / output | $1 / $5 | $2 / $10intro to Aug 31 · then $3 / $15 | $5 / $25fast mode $10 / $50 | $10 / $50 |
| Cache read (0.1×) | $0.10 | $0.20then $0.30 | $0.50 | $1.00 |
| Batch (50% off) | $0.50 / $2.50 | $1 / $5then $1.50 / $7.50 | $2.50 / $12.50 | $5 / $25 |
| Context window | 200k | 1Mstandard pricing across full window | 1Mstandard pricing across full window | 1Mstandard pricing across full window |
| Max output | 64k | 128k300k via Batch API beta | 128k300k via Batch API beta | 128k |
| Thinking | Extended | Adaptive | Adaptive | Adaptive · always on |
| Effort default | — | Highon API & Claude Code | Highon all surfaces | — |
| Latency | Fastest | Fast | Moderate | Slower |
| Tokenizer | Previous gen | New~30% more tokens per text | New~30% more tokens per text | New~30% more tokens per text |
| Reliable knowledge | Feb 2025 | Jan 2026 | Jan 2026 | Jan 2026 |
| Launched | Oct 2025 | Jun 30, 2026 | May 28, 2026 | Jun 9, 2026 |
Cache writes bill at 1.25× base input (5-minute TTL) or 2× (1-hour TTL); reads at 0.1×. Batch and caching discounts stack. Mythos 5 shares Fable 5’s specs and pricing but is invitation-only for approved organisations. Source: Anthropic models overview & pricing docs, July 4, 2026.
List price is only the opening bid. Output tokens cost 5× input on every tier, cache reads collapse input cost by 10×, and batch halves everything. Your effective rate is a function of workload shape, not the pricing page.
All four current models bill the full context window at standard per-token rates — a 900k-token request costs the same rate as a 9k one. Web search adds $10 per 1,000 searches on top of tokens.
Capability climbs to the top-right — and you pay for the top-right twice: in dollars and in seconds. The routing question is never “which model is best”, it’s “which point on this canvas does this task deserve”.
Capability tiers reflect Anthropic’s own positioning language; latency is Anthropic’s comparative rating. Blended = simple average of standard input and output list price. This is a positioning map, not a benchmark — see below for one of those.
Scores as disclosed around the May–June 2026 launches (Fable 5 figure is Anthropic-stated, not independently audited; Haiku 4.5 is not reported on this benchmark). Benchmarks are a prior, not a verdict — the only benchmark that settles routing is your own eval set on your own prompts.
This is the flow I use when designing routing policies: every workload enters at the top, and every “no” you can answer honestly is margin. The dashed rail is the part most teams forget — a cheap model plus a verifier plus an escalation path beats an expensive default almost everywhere.
Sticker prices don’t answer “what will my traffic cost”. Set your workload shape below — the maths runs on Anthropic’s published rates, including the cache-read discount and the new-tokenizer uplift most calculators ignore.
Monthly = 30 days. Assumes a warm cache (reads at 0.1× base input; write premiums and TTL churn excluded), no tool surcharges, list rates on the first-party API. The tokenizer uplift approximates Anthropic’s stated ~30% token increase on new-tokenizer models; real expansion varies by content. An estimator, not an invoice.
Notes from building model-routing infrastructure — the details that decide whether the ladder saves you money or quietly costs you it.
Whatever your platform falls back to when nobody chooses is your biggest AI cost line. Make the default an explicit, reviewed policy — Sonnet 5 today — not an accident of whoever wrote the first config.
Fable 5, Opus 4.8 and Sonnet 5 use a new tokenizer that yields ~30% more tokens for the same text (Anthropic-stated). Sticker comparisons against Haiku 4.5 or Sonnet 4.6 that ignore it are off by a third — Anthropic itself calls Sonnet 5’s intro price roughly cost-neutral versus 4.6 for exactly this reason.
Independent lab Artificial Analysis measured Sonnet 5 at ~$2.29 per task versus ~$1.99 for Opus 4.8 on its Intelligence Index at standard rates — the cheaper model spent more tokens and turns. Effort settings swing turn counts up to ~6×, a bigger cost lever than model choice. Measure cost per finished task.
At 0.1× reads, a warm 200k-token codebase costs $0.10 per pass on Opus 4.8 instead of $1.00. Structure prompts so the stable prefix caches — for agent loops, cache architecture matters more than model choice.
Sonnet 5’s $2/$10 window closes August 31, 2026; standard $3/$15 follows. Run your routing evals at both prices now so September’s invoice isn’t a surprise — and remember batch halves whichever rate applies.
You can’t reliably predict task difficulty up front, but you can verify outputs cheaply. Cheap-first routing plus a verifier plus escalation-on-failure turns the four-tier ladder into a cost optimiser — this is how I build routers, and it beats any single default.
The volume floor. If it runs a million times a day and no human reads it raw, it belongs here — escalate the exceptions.
The default. Most production traffic, most coding, most agents — especially while intro pricing holds.
The correctness tier. When wrong is expensive, the premium over Sonnet is the cheapest insurance you can buy.
The long horizon. Reserve it for autonomous, multi-hour work where quality compounds. A scalpel, not a default.