NezhaGateNezhaGate
简体中文English ✓Español한국어日本語
Log in Get API Key
Chat

Claude Fable 5

⧉
claude-fable-5

Claude Fable 5 belongs to the Anthropic Fable series: same generation and capability base as Opus 5 (adaptive thinking, tool use, image understanding, prompt caching), with a finer touch and richer emotional layering in Chinese narrative and long-form writing. Fully OpenAI-API compatible: set model to claude-fable-5; the native Anthropic Messages API is supported as well.

ChatWritingNarrativeImage understandingPrompt caching
Get API Key → API Docs

Live Test · Playground

Try out Claude Fable 5 right here (available after login).

Input

Advanced
Web search
Memory · multi-turn
Thinking

Conversation

Start a conversationType a message below to begin

About Claude Fable 5

Claude Fable 5 is a model in Anthropic's Fable series, same generation as Claude Opus 5 and built on the same capability base: adaptive thinking by default, native tool use, image input and prompt caching. It stands out on narrative and long-form writing - character motivation and emotional progression stay coherent and long pieces hold their structure - which suits fiction, scripts, brand copy and in-depth long-form articles. On NezhaGate it is available both through the OpenAI-compatible endpoint (set model to claude-fable-5) and through the native Anthropic Messages API. Billing is pay-as-you-go and failed requests are not charged.

Use cases

Fiction and screenwriting

Long-form narrative, character arcs and dialogue polish, staying consistent with characters and plot across many revisions.

In-depth long-form writing

Industry analysis, columns and reviews that need structure and flow, with more natural logical progression between paragraphs.

Brand and marketing copy

Product stories, brand narrative and ad scripts, with a steadier sense of emotional register than general-purpose models.

Rewriting and polishing

Expand, rewrite, restyle and translate existing drafts while preserving meaning and adjusting the voice.

General chat and code

It carries the same generation's reasoning and coding ability, so everyday tasks beyond writing work too.

How to choose

Choose Claude Fable 5 when content creation is the main job - narrative, long-form structure and emotional nuance are where it is strongest. If the workload is complex reasoning, large code projects or agent orchestration, its sibling Claude Opus 5 fits better and costs less; for high-frequency, low-cost everyday chat, Claude Sonnet 4.6 is more economical. All three share one endpoint - just change the model field.

FAQ

What is Claude Fable 5?
Claude Fable 5 is a model in Anthropic's Fable series, same generation as Claude Opus 5, sharing adaptive thinking, tool use, image input and prompt caching, and positioned toward narrative and long-form writing. On NezhaGate you can call it through the OpenAI-compatible API or the native Anthropic Messages API.
Fable 5 or Opus 5?
Same generation and the same capability base; they differ in emphasis and price. Fable 5 leans toward content creation with a finer touch for narrative and long-form writing, while Opus 5 leans toward complex reasoning and code engineering and costs less per token. Pick Fable 5 for stories and long articles, Opus 5 for engineering and agents.
Why is my temperature ignored?
Anthropic deprecated temperature / top_p / top_k on this generation and the upstream rejects them. The gateway strips those three parameters before forwarding, so the request succeeds with no error and sampling runs at the model default - existing OpenAI SDK code needs no changes. To steer the output, use reasoning_effort (low / medium / high / xhigh / max) to set thinking depth instead; all five rungs cost the same per token.
Does it support streaming and prompt caching?
Both. Set stream=true on the OpenAI-compatible endpoint for token-by-token output, or use the native endpoint for standard Anthropic SSE. Cache writes and cache hits are both verified working and are billed at their own cache-write / cache-read rates.
Will long articles get truncated?
That depends on max_tokens. With a small value a long piece stops midway (stop_reason max_tokens); raise max_tokens to let it finish in one call. Very large output budgets and inputs on the order of 200K tokens are both supported.

Related Models

Explore other models you can integrate.

View all →
Claude Opus 5 Chat

Claude Opus 5

claude-opus-5

Next-generation flagship from Anthropic

$4.00/1M in · $20.00/1M out View →
Claude Sonnet 4.6 Chat

Claude Sonnet 4.6

claude-sonnet-4-6

Balanced, efficient chat and coding model

$1.50/1M in · $7.50/1M out ↓50% View →