# Sonnet 5 vs Haiku 4.5 vs Opus 5: Which Claude Model Wins? (July 2026)
> Quick verdict: Sonnet 5 for most work, Haiku 4.5 when speed and cost matter, Opus 5 for hard code and long agentic chains. Head-to-head — Sonnet 5 vs Haiku 4.5, Opus 5 vs Sonnet 5, Opus 5 vs Opus 4.8 — with pricing, limits, and per-profession picks. Updated July 2026.
**Author:** [Alex Lowe](https://theaicareerlab.com/about) — Founder, The AI Career Lab
**Published:** 2026-05-20
**Updated:** 2026-07-27
**Last reviewed:** 2026-07-27
**Canonical URL:** https://theaicareerlab.com/blog/which-claude-model-should-you-use
**Category:** guide
**Tags:** Claude, Claude Opus 5, Claude Sonnet 5, Claude Haiku 4.5, Claude Opus 4.8, Claude Fable 5, model selection, 2026
---> **TL;DR.** **Sonnet 5** is the default Claude model and the right choice for most professional work. **Opus 5** — Anthropic's newest Opus, launched in July 2026 as the drop-in successor to Opus 4.8 at the same price — is the pick for the hardest code and long agentic chains. **Haiku 4.5** wins on speed and scale. **Fable 5** sits above them all when maximum capability justifies its separate billing.

Anthropic's current generally-available lineup as of late July 2026: **Sonnet 5** (the default), **Opus 5** (the new Opus-tier model — see our [Claude Opus 5 write-up](/blog/claude-opus-5-for-professionals-2026) for the launch details), **Haiku 4.5**, and **Fable 5**, a Mythos-class model that sits above the Opus tier (see our [Claude Fable 5 explainer](/blog/claude-fable-5-for-professionals) for its credits-based billing). The three workhorse tiers share a common feature surface (text + image input, vision, tool use, Skills, Projects) but differ meaningfully on capability, speed, cost, and context window. Picking the right one for the right task is still the highest-leverage tooling decision most professionals make in 2026.

This guide does the translation. Specs cited come from [Anthropic's model documentation](https://platform.claude.com/docs/en/about-claude/models/overview) as of July 2026.

## The three workhorse models at a glance

| | **Opus 5** | **Sonnet 5** | **Haiku 4.5** |
|---|---|---|---|
| **Position** | Strongest on hard code + agentic chains | The default; matches Opus-tier on most knowledge work | Fastest; near-frontier intelligence |
| **API ID** | `claude-opus-5` | `claude-sonnet-5` | `claude-haiku-4-5-20251001` |
| **Context window** | 1M tokens | 1M tokens | 200K tokens |
| **Max output** | 128K tokens | 128K tokens | 64K tokens |
| **Pricing (API)** | $5/M input, $25/M output | $2/$10 intro through Aug 31, 2026 — then $3/$15 | $1/M input, $5/M output |
| **Latency** | Moderate | Fast | Fastest |
| **Thinking** | Adaptive, effort-controlled (on by default) | Adaptive, effort-controlled (no manual extended-thinking mode) | Extended thinking (manual) |

Opus 5 keeps the effort-controlled adaptive thinking introduced with the recent Opus generations (`low` / `medium` / `high` / `xhigh` / `max`, defaulting to `high`) — and, unlike Opus 4.8, thinking is now on by default. Sonnet 5 uses the same effort system; the old Sonnet 4.6-era "Extended vs Adaptive thinking" distinction is gone.

## The single most important question: are you on claude.ai or the API?

**If you're on claude.ai** (the consumer/team chat product), Sonnet 5 has been the default for Free and Pro users since July 1, 2026 — and the default is genuinely the right model for most professional work. Opus 5 is available on all paid plans: it's the default model on Max and the strongest model available on Pro. Pro ($17/mo annual or $20/mo monthly) and above get the full picker. Switch freely as the task changes — there's no per-token cost to you, just usage limits per plan.

**If you're using the API** (or Claude Code, which uses the API tier you're on), each request costs by token. Sonnet 5's introductory pricing ($2/$10 through August 31, 2026) makes it the obvious default there too — with one caveat: its updated tokenizer produces 1.0–1.35x more tokens for the same content than Sonnet 4.6 did, so real-world costs run slightly higher than the headline rate suggests.

If you just want one default: **leave it on Sonnet 5**. The rest of this guide is about the exceptions.

## Pick Sonnet 5 for: almost everything (it's the default for a reason)

Sonnet 5 collapsed most of the old "when to pay for Opus" calculus. On real-world knowledge-work benchmarks it edged out Opus 4.8 (GDPval-AA v2: 1,618 vs 1,615), and it leads on computer-use tasks (OSWorld-Verified: 81.2%). In practice that means:

- **Client / borrower / customer communication** — the daily back-and-forth, at speed
- **Long-form structured documents** — memos, disclosure drafts, PRDs, case analyses. This used to be Opus territory; Sonnet 5 handles most of it at Opus-tier quality
- **Long-document analysis** — full case files, medical histories, board packages, policy documents. 1M context, same as Opus
- **Multi-step workflows** — Sonnet 5's agentic improvements were the headline of its release; most chained workflows don't need Opus
- **Iteration and polishing** — fast enough to think alongside you

### Profession-specific Sonnet 5 wins

- **Loan officers** running the four-audience pipeline update workflow daily across 8+ active loans
- **Real estate agents** generating listing descriptions, client emails, CMAs, market updates
- **Attorneys and paralegals** drafting memos and summaries from full source documents — verify the hardest analyses on Opus if the stakes demand it
- **Copywriters** and **community managers** doing iteration-heavy, voice-sensitive work
- **Most healthcare clinicians** drafting SOAP notes, treatment plans, patient education
- **Management consultants** synthesizing meeting notes into strategy memos
- **AI product managers** structuring specs and rollout plans

## Pick Opus 5 for: the hardest code and the longest agentic chains

Opus 5 is Anthropic's newest Opus-tier model — a step-change over Opus 4.8 on exactly the work Opus was already best at, delivered at the **same $5/$25 price**. Where it clearly earns the premium over Sonnet 5:

- **Complex software engineering** — multi-file features, larger refactors, end-to-end feature work in Claude Code. Opus 5's agentic-coding gains over Opus 4.8 are the headline of the release
- **The longest multi-step agentic workflows** — compliance automation, data-extraction pipelines, anything that chains dozens of steps where a single early error compounds. Opus 5 self-verifies its work as it goes
- **Second opinions on high-stakes deliverables** — when the document goes to a regulator, court, or board and you want the strongest Opus-tier model to draft or check it

Two practical notes. First, **Fast mode** on Opus 5 costs $10/M input and $50/M output for roughly 2.5x the response speed — so Opus depth with low latency carries a modest premium. Second, the `effort` parameter **defaults to `high`**; set it lower for faster, cheaper responses on simpler work.

### Profession-specific Opus 5 wins

- **Data scientists and technical operators** running hard multi-step pipelines in Claude Code
- **Healthcare compliance officers** running QSR gap audits across full DHF documents where miss-cost is extreme
- **AI compliance officers** producing pre-legal regulatory screens across EU AI Act tiers + Annex III + US state overlays
- **ESG sustainability analysts** mapping KPIs to multiple frameworks from a 200-page sustainability report

## Pick Haiku 4.5 for: speed and scale

Haiku 4.5 is the fastest current Claude model and the cheapest. It's also more capable than people expect — "near-frontier intelligence" per Anthropic's framing. The places it earns its place:

- **High-volume classification or extraction tasks** — running over thousands of inputs per day
- **Real-time chat surfaces** — when the response needs to feel instantaneous (in-product chatbots, customer support assistants)
- **Background AI features inside production applications** — the AI that runs invisibly behind a feature, where latency directly affects user experience
- **Cost-sensitive workflows where Sonnet's depth isn't worth the price** — at $1/$5, Haiku is the cheapest current model even against Sonnet 5's intro pricing

Haiku's constraints to know:
- **200K context window**, not 1M. Long documents need chunking
- **Manual extended thinking only** — no effort-controlled adaptive thinking
- **Reliable knowledge cutoff is Feb 2025** — older than Sonnet 5's (Jan 2026). For questions about events in 2025 or 2026, Haiku may have stale knowledge

### Profession-specific Haiku 4.5 wins

- **Customer support functions** building in-product AI chat where latency matters
- **Recruiters / HR** running high-volume resume classification or initial screening (with appropriate human-in-the-loop and EEOC-aware guardrails — see our [recruiter audit guide](/blog/ai-for-recruiters-guide))
- **Sales teams** generating high-volume personalized outreach where the template is the value-add and depth matters less
- **Internal tooling** — Slack bots, internal knowledge search, the AI behind ops dashboards
- **Real-time content moderation** in community management workflows

## Head-to-head: the matchups people actually search

### Sonnet 5 vs Haiku 4.5

**Verdict: Sonnet 5 for quality-sensitive work, Haiku 4.5 for speed- and volume-sensitive work.** If a human reads the output and judges it, use Sonnet 5; if the output feeds a pipeline, a chat widget, or a thousands-per-day loop, use Haiku 4.5.

The practical differences: Sonnet 5 has a 1M-token context window against Haiku's 200K, a newer knowledge cutoff (Jan 2026 vs Feb 2025), and effort-controlled adaptive thinking. Haiku answers faster and costs half of even Sonnet 5's intro rate ($1/$5 vs $2/$10). For most professionals working through claude.ai, this one is easy — Sonnet 5 is the default and the right call; Haiku is a builder's model.

### Opus 5 vs Sonnet 5

**Verdict: Sonnet 5 by default; Opus 5 when the task is genuinely hard technical work.** Sonnet 5 matches Opus-tier quality on most everyday knowledge work at less than half the price, so paying the Opus premium for drafting and analysis rarely makes sense.

Opus 5 pulls clearly ahead on complex software engineering, long agentic chains, and deep multi-step reasoning — the release's headline gains are exactly there. On claude.ai the calculus is even simpler: there's no per-token cost, so switch to Opus 5 whenever you hit hard code or a long agentic session and switch back after.

### Opus 5 vs Opus 4.8

**Verdict: Opus 5, unambiguously — it's better at the same price.** Opus 5 succeeds Opus 4.8 at the identical $5/$25 per MTok, with the same 1M context window and 128K max output, and delivers a substantial jump on agentic coding and long-horizon work.

Opus 4.8 remains available on the API and cloud platforms, so pinned integrations won't break — but there's no scenario where a new project should prefer it. API users: swap `claude-opus-4-8` for `claude-opus-5` and you get the upgrade at no price change. On claude.ai, Opus 5 has simply replaced Opus 4.8 as the Opus-tier option on paid plans. Full details in our [Opus 5 write-up](/blog/claude-opus-5-for-professionals-2026).

### Older matchups (previous-generation models)

These pairings still get asked about; the models remain available via the API even though newer options exist.

#### Opus 4.8 vs Sonnet 5

The pre-Opus-5 question. Sonnet 5 matched or beat Opus 4.8 on real-world knowledge work (GDPval-AA v2: 1,618 vs 1,615) while Opus 4.8 led on complex software engineering (SWE-bench Pro: 69.2% vs 63.2%). The answer was "Sonnet 5 unless it's hard code" — and with Opus 5 now out at the same price as 4.8, the hard-code answer has moved to Opus 5.

#### Opus 4.8 vs Sonnet 4.6

Both are previous-generation now. Opus 4.8 was the stronger model across the board — Sonnet 4.6's case was price. Today the same money buys strictly better options: Sonnet 5 costs less than Sonnet 4.6 did (through Aug 31) and beats it decisively, and Opus 5 replaces Opus 4.8 at the same price.

#### Haiku 4.5 vs Sonnet 4.6

Haiku 4.5 remains current; Sonnet 4.6 is legacy. If you're choosing between these two today, the real choice is Haiku 4.5 vs Sonnet 5 — Sonnet 5 is better and (through Aug 31, 2026) cheaper than Sonnet 4.6 ever was. Haiku still wins wherever latency and volume dominate.

## Where does Fable 5 fit?

Fable 5 is a **Mythos-class** model that sits above the Opus tier entirely (SWE-bench Pro: 80.3% vs Opus 4.8's 69.2%). It is Anthropic's most intelligent generally available model. The catch is billing: since July 20 it stays included (at reduced limits) only on Max and Team Premium plans; Pro and Team Standard access it via API credits at $10/$50 per million tokens, and API use requires accepting a 30-day data-retention term for safety monitoring.

The arrival of Opus 5 changes the escalation math: Opus 5 delivers close to Fable 5's frontier capability at half its API price, and on some agentic-coding benchmarks it even edges ahead. Practical rule: treat Opus 5 as the workhorse ceiling, and reserve Fable 5 for the sessions where the deepest reasoning is worth paying for separately. For the full story, billing tiers, and the data-retention caveat, see our [Claude Fable 5 explainer](/blog/claude-fable-5-for-professionals).

## The decision tree

If you only remember one decision rule, use this:

1. **Is it everyday professional work — drafting, analysis, client comms, documents, most workflows?** → **Sonnet 5**. The default is the right answer
2. **Is it genuinely hard code, or a long agentic chain where early errors compound?** → **Opus 5**
3. **Is latency the user-facing experience, or is this running at thousands-per-day volume?** → **Haiku 4.5**
4. **Is this the rare session where you want the most capable model available and your plan's Fable usage (or credits) covers it?** → **Fable 5**

## When to switch mid-conversation

On claude.ai you can switch models mid-conversation. The pattern that works for serious work:

- **Stay in Sonnet 5** for thinking, drafting, iterating — and for most final deliverables too
- **Switch to Opus 5** when you hit genuinely hard code or a long multi-step agentic task
- **Switch to Fable 5** for the occasional session where maximum capability is worth it and your plan allows

The old discipline — "draft in Sonnet, finish in Opus" — is mostly obsolete. Sonnet 5's knowledge-work quality means the finish-in-Opus step now only pays for itself on the hardest technical work.

## Effort levels: how Opus 5 and Sonnet 5 tune depth

Both Opus 5 and Sonnet 5 take an `effort` parameter that controls how much the model deliberates before answering — `low`, `medium`, `high` (the default), `xhigh`, and `max`. Lowering it trades depth for speed and cost; `xhigh` suits long-running agentic and coding sessions.

The question that used to come up constantly — "Opus on low effort vs Sonnet on high?" — has a cleaner answer now:

- **For knowledge work**, Sonnet 5 at default effort is simply the right tool. It benchmarks at Opus-tier level and costs less.
- **For hard code and long agentic chains**, Opus 5 wins on base capability — use it at `high` or `xhigh`. Notably, Opus 5's lower effort levels punch above their weight, so `medium` is a real cost lever on simpler technical work.
- **Rule of thumb:** pick the model by the task's *type* (knowledge work → Sonnet 5; hard technical work → Opus 5), then use `effort` to tune speed and cost within that model.

(Haiku 4.5 is the exception: it exposes manual Extended Thinking, which you invoke explicitly, rather than effort-controlled adaptive thinking.)

## What about the legacy models?

Anthropic still publishes documentation for Opus 4.8, Sonnet 4.6, Opus 4.7, Opus 4.6, Sonnet 4.5, Opus 4.5, Opus 4.1, and the original Claude 4 models. **Claude Sonnet 4 (`claude-sonnet-4-20250514`) and Claude Opus 4 (`claude-opus-4-20250514`) are deprecated and retired as of June 15, 2026.** **Opus 4.1 retires August 5, 2026** — migrate to `claude-opus-5`.

**Opus 4.8 is now the previous-generation Opus.** It remains active on the API and all cloud platforms, but moving to `claude-opus-5` is a same-price upgrade ($5/M input, $25/M output) with meaningfully better performance — the clearest model migration Anthropic has shipped in a while. **Sonnet 4.6 is a legacy model** too: if your integration pins `claude-sonnet-4-6`, update it to `claude-sonnet-5` — a better model at (through August 31) a lower rate, with the tokenizer caveat above.

## How API model versioning works

One subtle change starting with Claude 4.6: model IDs are pinned snapshots, not evergreen pointers. `claude-opus-4-8` won't auto-upgrade to Opus 5 — it stays on its snapshot, exactly as `claude-sonnet-4-6` stayed put when Sonnet 5 launched. This is a real change from earlier versioning patterns where some aliases moved.

For production integrations, this is the right behavior: you don't want your model silently changing under you. For staying current, it means actively migrating model strings when new versions ship — as everyone pinned to `claude-opus-4-8` or `claude-sonnet-4-6` should be doing right now.

## Pricing context

Per-token pricing as of July 2026 from [Anthropic's documentation](https://platform.claude.com/docs/en/about-claude/pricing):

- **Opus 5:** $5/M input, $25/M output (Fast mode: $10/$50) — unchanged from Opus 4.8
- **Sonnet 5:** introductory $2/M input, $10/M output through August 31, 2026; standard $3/$15 from September 1. Updated tokenizer produces 1.0–1.35x more tokens for the same content vs Sonnet 4.6
- **Haiku 4.5:** $1/M input, $5/M output

For most professionals working through claude.ai rather than the API, per-token pricing is academic — usage limits are per-plan, not per-token. But if you're using Claude Code (which uses your API tier), or you're integrating Claude into your own product, the per-token pricing is where the per-model cost calculus lives.

[Anthropic's pricing page](https://claude.com/pricing) has the current consumer plan details. Verify before committing to a plan.

## Bottom line

Leave claude.ai on Sonnet 5 — the default is the right answer for most professional work. Reach for Opus 5 when the task is genuinely hard code or a long agentic chain, Haiku 4.5 when speed or volume is the point, and Fable 5 for the rare session where maximum capability justifies its separate billing.

The era of "just use Opus for everything" is over — Sonnet 5 matches the Opus tier on most knowledge work at a fraction of the cost. Pick the model for the task, not the task for the model.

For the deeper dives, see our [Claude Opus 5 write-up](/blog/claude-opus-5-for-professionals-2026), [Claude Fable 5 explainer](/blog/claude-fable-5-for-professionals), and the [Claude Sonnet 5 launch write-up](/blog/claude-sonnet-5-for-professionals-2026). For Claude vs ChatGPT, see our [profession-specific comparison hub](/compare).

---

*This article cites model specifications as published in [Anthropic's model documentation](https://platform.claude.com/docs/en/about-claude/models/overview) and pricing as published at [claude.com/pricing](https://claude.com/pricing) as of July 2026. Anthropic updates model availability, capabilities, and pricing frequently. Verify current state before procurement or integration decisions.*
## Frequently asked questions

### Is Opus 5 better than Sonnet 5?

For hard code and long agentic chains, yes — that's exactly what Opus 5 is built for, and it's a step-change over Opus 4.8 in that territory. For everyday knowledge work — drafting, analysis, client communication — Sonnet 5 remains the right default: it matches Opus-tier quality on most of those tasks at a lower price ($2/$10 intro API pricing vs Opus 5's $5/$25). Default to Sonnet 5; escalate to Opus 5 when the task is genuinely hard technical work.

### What is the difference between Opus 5, Sonnet 5, and Haiku 4.5?

Opus 5 is Anthropic's newest and most capable Opus model (1M-token context, strongest on deep reasoning, hard code, and long agentic work). Sonnet 5 is the default workhorse (1M context, 128K output, matches Opus-tier quality on most knowledge work, cheaper). Haiku 4.5 is the fastest and cheapest (200K context, best for high-volume and real-time tasks). API pricing per million input/output tokens: Opus 5 $5/$25, Sonnet 5 $2/$10 intro through Aug 31 2026 (then $3/$15), Haiku $1/$5.

### Sonnet 5 vs Haiku 4.5 — which should I use?

Sonnet 5 for anything where quality matters: documents, analysis, drafting, multi-step workflows. Haiku 4.5 when speed or volume is the point: real-time chat surfaces, high-volume classification, background AI features where latency directly affects user experience. Haiku is also the cheapest option at $1/$5 per million tokens, but it has a smaller 200K context window and an older knowledge cutoff (Feb 2025), so long documents and current-events questions belong on Sonnet 5.

### Is Opus 5 better than Opus 4.8?

Yes — and it costs the same. Opus 5 succeeds Opus 4.8 at the identical $5/$25 per-million-token API price, with the same 1M context window, and delivers a substantial jump on agentic coding and long-horizon work. If you're on claude.ai, Opus 5 simply replaces Opus 4.8 as your Opus-tier option. If you pinned claude-opus-4-8 in the API, updating to claude-opus-5 is effectively a free capability upgrade.

### Which Claude model should I use on claude.ai?

Leave the model on Sonnet 5 — it's the default for Free and Pro for a reason. Switch to Opus 5 for genuinely hard code or long multi-step agentic work; it's available on all paid plans and is the default on Max. Pro ($20/month, or $17/month billed annually) and above unlock the full picker, and there is no per-token cost on claude.ai — just per-plan usage limits — so switching costs you nothing when the task warrants it.

### How much does each Claude model cost?

API pricing as of July 2026: Opus 5 — $5/M input, $25/M output (same as Opus 4.8; Fast mode $10/$50); Sonnet 5 — introductory $2/M input, $10/M output through August 31, 2026, then $3/$15 (note: Sonnet 5's updated tokenizer produces 1.0–1.35x more tokens for the same content); Haiku 4.5 — $1/M input, $5/M output. On claude.ai you pay per plan, not per token: Free (Sonnet 5 plus limited Haiku), Pro $20/month ($17 annual), Max $100–$200/month, Team $25–$125/seat, Enterprise custom.

### What happened to Claude Fable 5 — can I use it now?

Yes, with caveats. Fable 5 launched June 9, 2026, was suspended June 12 under a US government export-control directive, and was restored globally on July 1. Since July 20 it stays included (at reduced limits) only on Max and Team Premium; Pro and Team Standard users access it via API credits at $10/$50 per million tokens. It's Anthropic's most intelligent generally available model — a Mythos-class tier above Opus. See our Fable 5 explainer for the full story and billing details.

---

*Canonical version: https://theaicareerlab.com/blog/which-claude-model-should-you-use*
*This document is the Markdown companion served for AI crawlers and answer engines. See the canonical URL for the rendered version with navigation, related content, and interactive elements.*