As of November 9, 2026, a US Claude Pro subscription gives you access to Opus, Sonnet, and Haiku models in Claude’s apps. Fable is also available to Pro subscribers, but its use is billed through separate usage credits rather than included in Pro’s plan limits. Pro does not include Claude API usage. For production apps, API model access and per-token billing are separate from the subscription. (Anthropic plans and pricing; Fable models on your plan)
Which models can I use with Claude Pro?
Anthropic’s plan comparison lists Opus, Sonnet, and Haiku for Pro subscribers. Fable is listed separately: it can be used with separately billed usage credits, not the plan’s included usage allowance. Pro access is subject to plan usage limits; check Anthropic’s current plan details for the terms that apply to your account. (Anthropic plans and pricing; Fable models on your plan)
Pro is not an API plan. Claude API usage through the Claude Console is billed separately from a paid Claude subscription. Don’t use API token rates to estimate the subscription cost, or assume that a model available through the API is included in Pro. (What is the Pro plan?; Why API usage is billed separately)
Claude API models for production
The API model overview lists Haiku, Sonnet, Opus, and Fable, with distinct model IDs and workload descriptions. Treat those descriptions as starting points, not as a guarantee that a model will perform best on your tasks. (Anthropic’s model overview)
| API model | API ID | Starting point by workload |
|---|---|---|
| Claude Haiku 5.5 | claude-haiku-5-5 |
Evaluate for high-volume or latency-sensitive tasks such as classification, extraction, and routing. |
| Claude Sonnet 5.5 | claude-sonnet-5-5 |
Consider for work where a balance of speed and capability matters. |
| Claude Opus 5.5 | claude-opus-5-5 |
Evaluate for demanding reasoning, coding, and longer-running agent work. |
| Claude Fable 5.1 | claude-fable-5-1 |
Test for especially demanding tasks when evaluations show a need for it. |
API pricing: reported figures, not confirmed current rates
The following USD-per-million-token API rates are reported by third-party model guides, not confirmed current Anthropic rates. For Haiku, those guides report different input and output rates depending on whether the prompt is up to or over 100,000 tokens. Confirm current rates and applicable conditions on Anthropic’s API pricing page before budgeting. These are API rates—not Claude Pro subscription prices. (Third-party model guide; third-party Sonnet listing)
| API model | Reported API price per 1 million tokens, input / output (USD) |
|---|---|
| Claude Haiku 5.5 | $0.10 / $0.50 for prompts up to 100,000 tokens; $0.50 / $2.50 above that threshold |
| Claude Sonnet 5.5 | $2 / $10 |
| Claude Opus 5.5 | $4 / $20 |
| Claude Fable 5.1 | $10 / $50 |
How to choose a production model
- High-volume or bounded tasks: Start by evaluating Haiku for classification, extraction, or routing. If prompts are long, confirm the applicable price before deciding on cost.
- General product tasks: Include Sonnet as a candidate, then judge its fit against your quality and latency requirements.
- More demanding reasoning or coding: Test Opus when your evaluation tasks require it, and compare any quality gain with its cost and latency in your environment.
- The most demanding tasks: Consider Fable only if evaluation shows that it provides enough value to justify its separate usage-credit billing for Pro subscribers or its API cost for API use.
A practical comparison is to run the same representative prompts and success criteria through two or more candidate API models. For example, a code-review service could evaluate Haiku for simple triage, then compare Sonnet and Opus on reviews requiring deeper reasoning. This is an illustrative evaluation plan, not a claim that one model has been tested or wins universally. Track quality, latency, and API cost per successful task; keep model IDs fixed during the comparison so results are reproducible.
Bottom line
If you mean Claude Pro, Opus, Sonnet, and Haiku are listed for the subscription; Fable uses separately billed usage credits. Pro usage limits apply. If you mean production API models, compare Haiku, Sonnet, Opus, and—when your evaluation justifies it—Fable. API access and billing are separate from Pro, and the best choice depends on your workload. Treat the API rates above as third-party reported figures until confirmed against Anthropic’s current pricing page.
Sources
Plans & pricing | Claude by Anthropic
Claude Fable models on your plan | Claude Help Center
What is the Pro plan? | Claude Help Center
Why is Claude API usage billed separately from my paid Claude plan? | Claude Help Center
Models overview | Claude Platform Docs