Claude Code fast mode cost: Fable 5.2 at $16/$80 and the Opus 4.8 cutoff on November 30
Fast mode for Fable 5.2 costs $16 and $80 per MTok for up to 2.5x output speed. Opus 4.8 fast mode returns errors after November 30. Here is what to decide.
Claude Code fast mode on Fable 5.2 costs $16 per MTok of input and $80 per MTok of output, and Anthropic says it delivers up to 2.5x higher output tokens per second. Fast mode for Opus 4.8 was deprecated on October 30, and requests with speed: "fast" return an error after November 30. Anthropic names Opus 5.5 as the migration target.
TL;DR: Fast mode on Fable 5.2 launched November 9 at $16 input and $80 output per MTok, with up to 2.5x higher output tokens per second. Opus 5.5 fast mode is listed at $8 and $40. If you still run Opus 4.8 with fast mode on, the API rejects those requests after November 30, so switch to Opus 5.5 with/modelor run/fast off.
Key Takeaways
- Fable 5.2 price: fast mode is $16 input and $80 output per MTok, against $8 and $40 at standard speed.
- Speed claim: Anthropic says up to 2.5x higher output tokens per second, and the gain is on output speed, not time to first token.
- Deadline: Opus 4.8 fast mode stops working on November 30, and the failure is an error.
- Billing path: on subscription plans, fast mode draws from usage credits only.
- Cache cost: turning fast mode on mid-conversation bills the whole context at the full fast uncached input price.
What does Claude Code fast mode cost on Fable 5.2?
Fast mode on Fable 5.2 costs $16 per MTok of input and $80 per MTok of output. The Claude Platform release notes list the launch under November 9 and say fast mode "delivers up to 2.5x higher output tokens per second." Standard Fable 5.2 pricing is $8 and $40 per MTok on the Anthropic pricing page, so by simple arithmetic ($16 against $8, $80 against $40) the fast rate is twice the standard rate for both input and output; no page states that ratio.
This table compares the two modes.
| Item | Standard Fable 5.2 | Fast mode on Fable 5.2 | Source |
|---|---|---|---|
| Input per MTok | $8 | $16 | pricing page; Claude Code fast mode docs |
| Output per MTok | $40 | $80 | pricing page; Claude Code fast mode docs |
| Speed | Not stated | Up to 2.5x higher output tokens per second | release notes |
| Cache | Separate cache from fast mode | Full context billed uncached at the fast input price when enabled mid-conversation | Claude Code fast mode docs |
The speed figure is Anthropic's own claim and says "up to," so measure it on your own prompts. The Claude Code fast mode docs list every supported model:
| Model | Fast input (per MTok) | Fast output (per MTok) | Status in the docs |
|---|---|---|---|
| Fable 5.2 | $16 | $80 | Available |
| Opus 5.5 | $8 | $40 | Available |
| Opus 5 | $10 | $50 | Research preview |
| Opus 4.8 | $10 | $50 | Deprecated, removal scheduled for November 30 |
What happens to Opus 4.8 fast mode on November 30?
After November 30, any request to claude-opus-4-8 with speed: "fast" returns an error. The October 30 release notes entry says so and tells you to migrate to Claude Opus 5.5 to keep fast mode.
The Claude Code docs describe the same cutoff from the CLI side: if Opus 4.8 is still your model with fast mode on after that date, "the API rejects your requests." There are two fixes. Run /model and switch to Opus 5.5, which needs Claude Code v2.1.280 or later, and fast mode stays on. Or run /fast off to stay on Opus 4.8 at standard speed.
There is precedent for a hard failure. When fast mode for Opus 4.7 was removed on July 24, the release notes say requests returned an error and, unlike Opus 4.6, did not fall back to standard speed.

How do you turn on fast mode in Claude Code?
You turn on fast mode in the CLI by running /fast, pressing Space to toggle, and pressing Enter to confirm. Fast mode on Fable 5.2 requires Claude Code v2.1.321 or later, and the docs warn that setting "fastMode": true alone has no effect on Fable 5.2 until you run /fast while using it.
Your plan decides the billing path. Per the docs, subscription users on Pro, Max, Ultra, Team and Enterprise get fast mode through usage credits only, not through subscription rate limits. Team and Enterprise organizations also need an Owner to enable it first. Console organizations pay per token with the rest of their API usage.
How does fast mode change the bill in a long session?
The first time you enable fast mode in a conversation, you pay the full fast mode uncached input token price for the entire conversation context. The Claude Code docs say the charge is smallest when you enable it at the start, and that fast mode and standard speed keep separate caches, so a later switch can read part or all of the context uncached again.
Rate limits add a second path. When you hit the fast mode rate limit, the docs say fast mode falls back to standard speed, you continue at standard pricing, and it re-enables after the cooldown. The fast mode API page adds that the fallback causes a prompt cache miss, because requests at different speeds do not share cached prefixes.

When is fast mode worth paying for?
Fast mode is worth paying for when a person is waiting on each reply. The Claude Code docs recommend it for interactive work such as rapid iteration and live debugging, and recommend standard speed for long autonomous tasks, batch processing, CI/CD pipelines and cost-sensitive workloads.
The API page says fast mode runs the same model with a faster inference configuration, and that the speed benefit is on output tokens per second, not time to first token. It is also unavailable with the Batch API and with a Priority Tier commitment.
Frequently Asked Questions
How much does Claude Code fast mode cost?
Fast mode costs $16 per MTok of input and $80 per MTok of output on Fable 5.2, $8 and $40 on Opus 5.5, and $10 and $50 on Opus 5 and Opus 4.8, according to the Claude Code docs. On subscription plans the spend comes out of usage credits.
Does Opus 4.8 fast mode fall back to standard speed after November 30?
The release notes say requests to claude-opus-4-8 with speed: "fast" will return an error after removal. They do not describe a fallback for Opus 4.8. Switch to Opus 5.5 with /model to keep fast mode, or run /fast off.
Is fast mode a different model with different quality?
No. The API docs say fast mode runs the same model with a faster inference configuration, with no change to intelligence or capabilities.
Conclusion
If you run Opus 4.8 with fast mode on, move to Opus 5.5 or turn fast mode off before November 30. If you are weighing Fable 5.2 fast mode at $16 and $80 per MTok, turn it on at the start of a session for interactive work and keep it off for batch and CI jobs.
Learn from me

Claude Code in Practice, my Maven cohort. Master Claude Code from fundamentals to advanced orchestration: skills, subagents, hooks, MCP, and production automation. Join the next cohort →
Hire us
Traversaal.ai. We're a team of forward deployed engineers solving the toughest AI problems for Fortune 100 companies: document intelligence, agentic data platforms, and real-time web intelligence, deployed in production. Work with our team to deploy your next agentic ecosystem. Talk to Traversaal.ai →
Join us
Want to solve these problems with us? We're always looking for forward deployed engineers who want to ship production AI. jobs@traversaal.ai