Official commercial channel
Claude Fable 5 API
Anthropic’s tier above Opus — the one you pick when the ceiling is the point.
Context
1M
Max output
128K
Serving precision
Vendor-managed
Licence
Proprietary
Claude Fable 5 is the generally available model in the Mythos class Anthropic introduced above Opus — the restricted Claude Mythos 5 in the same class is offered only to a limited set of cyber defenders. Its 1-million-token window is the API default rather than an opt-in beta, with no long-context price premium, and output tops out at 128,000 tokens across text, image and file input. Adaptive thinking is always on, raw thinking is never returned, and effort, task budgets, memory, context editing and compaction all shipped at launch. Vendor list pricing is $10 per million input and $50 per million output.
Proprietary, provided through the vendor’s official commercial licensing. Contract, invoice, audit trail — not an account pool and not a reverse-engineered client protocol.
Specifications
- Model ID
- claude-fable-5
- Published by
- Anthropic
- How it is served
- Official commercial channel
- Licence
- Proprietary
- Context
- 1M
- Max output
- 128K
- Serving precision
- Vendor-managed
- Input modalities
- Text · Image
- Released
- 2026-06-01
- Price
- See pricing
What it is good at
- A tier above Opus
- Mythos is a class Anthropic created above Opus, aimed at autonomous knowledge work and coding. If the hardest judgement calls are the job and quality is worth paying for, it deserves one honest comparison on your own tasks.
- A million tokens with no premium
- The 1M window is the API default, not something to opt into, and long context is not priced separately. One fewer variable when you estimate the cost of very long inputs.
- Task budgets and context compaction
- Effort, task budgets, memory, context editing and compaction were all available at launch. On long autonomous runs these are the practical levers for controlling both cost and drift.
Where it fits
- The hardest reasoning and judgement work you have
- Long-running autonomous knowledge-work and coding agents
- Capability and cost comparisons against Opus 5
- Deep analysis and drafting over million-token documents
Three steps
Change one line of base_url, put the model ID below in the `model` field, and leave the rest of your code alone.
model: "claude-fable-5"
curl https://api.nexaix.net/v1/chat/completions \
-H "Authorization: Bearer $NEXAIX_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "claude-fable-5",
"stream": true,
"messages": [
{ "role": "user", "content": "Summarise this contract." }
]
}'import os
from openai import OpenAI
client = OpenAI(
# the only new line
base_url="https://api.nexaix.net/v1",
api_key=os.environ["NEXAIX_API_KEY"],
)
# everything below is your existing code
stream = client.chat.completions.create(
model="claude-fable-5",
stream=True,
messages=[
{"role": "user", "content": "Summarise this contract."}
],
)
for chunk in stream:
print(chunk.choices[0].delta.content or "", end="")import OpenAI from "openai";
const client = new OpenAI({
baseURL: "https://api.nexaix.net/v1", // the only new line
apiKey: process.env.NEXAIX_API_KEY,
});
// everything below is your existing code
const stream = await client.chat.completions.create({
model: "claude-fable-5",
stream: true,
messages: [
{ role: "user", content: "Summarise this contract." }
],
});
for await (const chunk of stream) {
process.stdout.write(chunk.choices[0]?.delta?.content ?? "");
}About this model
Fable 5 or Opus 5?
Usually Opus 5: it lists at half the price and lands within 0.5% of Fable 5’s peak on CursorBench at half the cost per task. Reach for Fable 5 where you have confirmed you need that last increment of ceiling. Run both with the same key against your real tasks first.
Any difference in data retention versus Opus 5?
Yes, and we would rather say it plainly: Opus 5 supports zero data retention while Fable 5 currently carries a 30-day retention requirement. Nothing is written to disk on our side for either, but the upstream difference is real. If you have hard compliance requirements, prefer Opus 5.
Do you offer Claude Mythos 5?
No. Per Anthropic, Mythos 5 is offered only to a restricted set of cyber defenders and is not generally available.
Also worth a look
- Official commercial channelClaude Opus 5A million tokens that never degrade, five-level effort control, zero data retention.claude-opus-5
- Official commercial channelGPT-5.6 SolThe frontier tier of the GPT-5.6 family: 1.05M context, built for long-horizon agents.gpt-5.6-sol
- Official commercial channelGPT-5.6 TerraThe everyday tier: the same window as Sol, at half the list price.gpt-5.6-terra
Run your own eval on it
Take the trial credit, run the evals you already have against this model, and then decide whether it gets your production traffic.
