Where do I create an API key and buy tokens?

Both happen on api.reapi.ai — the chat workspace runs as its own platform separate from the image / video task gateway at reapi.ai. Sign up at api.reapi.ai, generate a key under API Keys, and top up under Top Up. A reapi.ai/settings/apikeys key will not authenticate against the chat endpoint.

Is Claude Fable 5 OpenAI-compatible?

Yes. Claude Fable 5 is a drop-in for OpenAI's /v1/chat/completions — same request shape, same `messages` array, same `stream` and `max_tokens` parameters, same SSE wire format. Most teams migrate by changing the base URL to https://api.reapi.ai/v1, swapping the API key, and setting `model: "claude-fable-5"`. The native Anthropic Messages format at /v1/messages is also available. Note that Claude Fable 5 takes no sampling parameters — leave `temperature` and `top_p` out of the request.

What is the Claude Fable 5 context window?

Claude Fable 5 has a 1M token context window by default and supports up to 128K output tokens per response on the synchronous API. Long inputs benefit from prompt caching — cache stable parts of the prompt once and pay only the cache-read rate on subsequent calls.

How is Claude Fable 5 different from Claude Opus 4.8?

Claude Opus 4.8 is Anthropic's most capable Opus-tier model; Claude Fable 5 sits above it as the most capable model Anthropic widely releases, built for the most demanding reasoning and long-horizon agentic work. Claude Fable 5 also keeps adaptive thinking always on and adds structured refusals with automatic fallback. Both offer a 1M context window, 128K max output, and vision input on the same api.reapi.ai key.

How does adaptive thinking work on Claude Fable 5?

Adaptive thinking is the only thinking mode on Claude Fable 5 — it applies on every call and cannot be disabled. Use the effort parameter (low through max) to control how much reasoning Claude Fable 5 invests per request. Raw chain-of-thought is never returned; summarized thinking is available on the native /v1/messages surface.

What happens when Claude Fable 5 declines a request?

Claude Fable 5 includes safety classifiers that can refuse certain requests. A refusal comes back as a successful response with a structured refusal stop reason — not an HTTP error — and a request refused before any output is generated is not billed. A fallback mechanism can retry a refused request on another Claude model, and routing the retry yourself on the same api.reapi.ai key works too.

Does Claude Fable 5 support vision and multimodal input?

Yes. Claude Fable 5 accepts text and image inputs in the same call — useful for screenshots, diagrams, document scans, chart analysis, and multimodal review flows. Images are passed as URL content parts on the standard Chat Completions payload.

Claude Fable 5 — Anthropic's Most Capable Model

Claude Fable 5 is Anthropic's most capable widely released model for the most demanding reasoning and long-horizon agentic coding — 1M context, 128K max output, always-on adaptive thinking, vision input, and OpenAI-compatible /v1/chat/completions in one call. Pay-as-you-go pricing in USD.

Claude Fable 5modelclaude-fable-5

Claude Fable 5 playground

Open the chat playground to run Claude Fable 5 through the OpenAI-compatible chat completions surface with your api.reapi.ai key.

Open chat playground

What you can build with this model

Real-world workflows and production use cases you can build and ship with this model.

Frontier long-horizon agentic coding

Claude Fable 5 is built for the most demanding agentic work Anthropic's models can do — week-long refactors, codebase-scale migrations, and autonomous runs that must stay coherent across hundreds of steps. Where lighter models drift, Claude Fable 5 holds the plan, verifies its own output, and keeps going.

Read the API docs

Claude Fable 5 reasoning through a hard multi-step problem

The hardest reasoning problems

When a problem sits past what Opus-tier models handle cleanly — deep multi-step reasoning, ambiguous specifications, high-stakes analysis — Claude Fable 5 is the model Anthropic positions above everything else it widely ships. Adaptive thinking is always on, and the effort parameter scales how much reasoning Claude Fable 5 invests in each call, up to max.

Claude Fable 5 reasoning across a 1M-token analysis pack

Large-context analysis and review

Feed entire codebases, long research packs, multi-file technical material, or full policy documents into a single Claude Fable 5 request. The 1M token context window is billed at standard per-token rates — no long-context premium — so analysis-heavy Claude Fable 5 workflows rarely need chunking.

Pricing

Credit-based — 1 credit = $0.001 USD. Pay only for completed generations.

Category	Unit	Price
Tokens
Input tokens	1M tokens	$10
Output tokens	1M tokens	$50
Cache read	1M tokens	$1
Tools
Web search	request	$0.015

Why reAPI

OpenAI-compatible drop-in for Anthropic's top tier

Claude Fable 5 speaks OpenAI Chat Completions verbatim. Moving an existing OpenAI integration to Claude Fable 5 is a base URL, an API key, and a model-string change — not a platform rewrite. The same `messages` array, the same streaming format, and the native Anthropic /v1/messages surface is available too for SDK callers that prefer it.

The most capable model Anthropic widely ships

Claude Fable 5 sits a full tier above Claude Opus 4.8 — Anthropic's most capable widely released model, generally available since June 9, 2026. Route the hardest reasoning and the longest agentic runs to Claude Fable 5; send everyday premium coding to Opus-tier models on the same key.

One key across GPT, Claude, and Gemini

A single api.reapi.ai key unlocks Claude Fable 5 alongside GPT-5.5, Claude Opus 4.8, Gemini 3.1 Pro, and every other frontier chat model on the platform. Compare vendors, add fallbacks, and route traffic per call with a configuration change instead of an integration project.

Claude Fable 5 vs Claude Opus 4.8

Claude Fable 5 is the tier above Opus — the most capable model Anthropic widely releases, launched June 9, 2026. Here is how it compares with Claude Opus 4.8, Anthropic's strongest Opus-tier model, based on Anthropic's own documentation.

Capability

Claude Fable 5 on reAPI

Claude Opus 4.8

Capability tier

Anthropic's most capable widely released model — built for the most demanding reasoning and long-horizon agentic work.

Anthropic's most capable Opus-tier model for complex reasoning and agentic coding.

Adaptive thinking

Always on — the only thinking mode on Claude Fable 5; depth tuned with the effort parameter up to max.

Adaptive thinking supported but optional; runs without thinking when unset.

Refusals and fallback

Safety classifiers return structured refusals; a fallback mechanism retries on another Claude model automatically.

Standard request handling without the classifier-based refusal and fallback flow.

Refusal billing

Requests refused before any output is generated are not billed; fallback credit offsets the prompt-cache cost of switching.

Standard billing on all completed requests.

Pricing tier

Premium per-token rates reflecting the top capability tier — see the live pricing card above.

Lower per-token rates on the Opus tier.

Context, output & vision

1M-token context, 128K max output, text + image input.

1M-token context, 128K max output, text + image input — at parity.

Comparison reflects publicly documented behavior from Anthropic's Claude Fable 5 launch documentation and model overview at the time of writing. Model behavior and pricing can change; check the pricing card above and the API docs for current values.

Ship Claude Fable 5 in three steps

step 01
Create an account and key on api.reapi.ai
Sign up at api.reapi.ai, open the console, generate an API key under API Keys, and top up tokens under Top Up. The chat workspace is separate from the reapi.ai image/video gateway — keys do not cross over.
Open
step 02
Send your first Claude Fable 5 request
POST https://api.reapi.ai/v1/chat/completions with `model: "claude-fable-5"`, your `messages` array, and `max_tokens` set generously. The endpoint is OpenAI-compatible, including streamed responses; the native Anthropic /v1/messages format works too. Claude Fable 5 takes no sampling parameters — leave `temperature` and `top_p` out of the request.
Open
step 03
Tune for cost and stability
Use prompt caching for stable system prompts and recurring long inputs to bring repeated-context costs down. Reserve Claude Fable 5 for the highest-value calls and route everything else to a cheaper model on the same key.
Open

Frequently asked questions

Common questions about this model.

Claude Fable 5 is billed pay-as-you-go in USD against your api.reapi.ai token balance — see the pricing card on this page for the live per-1M-token input and output rates, the prompt-caching cache-read rate, and the per-request web-search rate. The full 1M context window is billed at standard rates with no long-context premium, and requests Claude Fable 5 refuses before generating output are not billed.

Related models

Explore more models in the same category.

View all models

Chat

Anthropic

Claude Opus 4.8

Anthropic's Claude Opus 4.8 — 1M context, 128K output, most-capable reasoning and agentic coding.

From $5.00 per 1M tokens

Chat

Anthropic

Claude Opus 4.7

Anthropic's Claude Opus 4.7 — 1M context, 128K output, premium coding and agent reasoning.

From $2.00 per 1M tokens

Chat

Anthropic

Claude Sonnet 4.6

Anthropic's Claude Sonnet 4.6 — balanced quality and speed for everyday production chat, code review, and mid-complexity agents.

From $2.00 per 1M tokens

Chat

MiniMax

MiniMax M3

MiniMax M3 — open-weight frontier coding, 1M context, native thinking and multimodal input, at a low per-token price.

From $0.600 per 1M tokens

View all models

start building

Ready to ship?

Try it in the playground or grab an API key to integrate now.

Get API key View API docs

What you can build with this model

Real-world workflows and production use cases you can build and ship with this model.