Home / AI Model & Price Tracker / Week of Nov 9–15, 2026

AI Model & Price Tracker · Week ending November 15, 2026

AI Price Tracker, Nov 9–15: Claude’s Top Model Drops 20%, Grok Holds Steady

One flagship price cut, one new premium speed tier, and zero changes to consumer plans. Here is the full table you can budget from, plus the deadline coming on Nov 30.

The short answer

In the week of Nov 9–15, 2026, the only list-price change across Grok and Claude was Claude Fable 5.2, which launched at $8 per million input tokens and $40 per million output tokens (Fable 5.1 cost $10 and $50). Fable 5.2 also got a fast mode at $16 and $80. Grok API prices and both companies’ consumer plans were unchanged. The next deadline is Nov 30, when Claude Opus 4.8 fast mode is removed.

1list-price change this week
$0.10cheapest input per million tokens (Claude Haiku 5.5, prompts ≤100k)
Nov 30next retirement deadline

Key takeaways

  • New: Claude Fable 5.2 at $8 input and $40 output per million tokens, with $0.20 cache reads (Nov 9).
  • New: Fable 5.2 fast mode at $16 input and $80 output per million tokens, for up to 2.5x output speed.
  • Unchanged: grok-4.7 at $2.00 input and $6.00 output per million tokens (prompts under 200k).
  • Unchanged: The X Subscription (Plus $8, Premium $30, Super $100, Ultra $200 a month) and Claude plans (Pro $20, Max $100, Ultra $200 a month).
  • Deadline: Claude Opus 4.8 fast mode is removed on Nov 30, 2026.
Free to read and cite. The weekly tables below are public. Link back if you use them. The analysis and the “Do this next” list are for subscribers.

What changed, Nov 9–15

Date Vendor Item Before After
Nov 9 Anthropic Claude Fable 5.2 (new; succeeds 5.1 as default Fable) Fable 5.1: $10 / $50 $8 / $40
Nov 9 Anthropic Fable 5.2 cache reads 5.1: $0.25 $0.20
Nov 9 Anthropic Fable 5.2 fast mode (new) — $16 / $80
Nov 9 Anthropic Claude Code default Fable model Fable 5.1 Fable 5.2

Prices are per million tokens, input / output. Fable 5.1 remains available at its old price. If your code names claude-fable-5-1, you will not get the lower price automatically. You have to change the model name.

Not changed: every Grok API model, every Grok Imagine and Grok Voice price, The X Subscription, and all Claude consumer and team plans.

The current API price table

The table is sorted by output price, because output usually drives the bill in chat and agent work. All prices are per million tokens for standard-length prompts.

Model Input Cached input / cache read Output Context
Claude Haiku 5.5 (≤100k prompt) $0.10 $0.01 $0.50 —
grok-build-0.1 $1.00 $0.20 $2.00 256k
grok-4.3 $1.25 $0.20 $2.50 1M
Claude Haiku 4.5 $1.00 $0.10 $5.00 —
grok-4.7 $2.00 $0.50 $6.00 500k
Claude Sonnet 5.5 $2.00 $0.10 $10.00 —
Claude Opus 5.5 $4.00 $0.20 $20.00 —
Claude Opus 4.8 $5.00 $0.50 $25.00 —
Claude Fable 5.2 $8.00 $0.20 $40.00 1M
Claude Fable 5.1 $10.00 $0.25 $50.00 —

“—” means the context size was not listed beside the price on the vendor’s pricing page. We do not fill gaps with guesses.

Output price per million tokens, Nov 15, 2026 (USD)Claude Haiku 5.5$0.50grok-build-0.1$2.00grok-4.3$2.50Claude Haiku 4.5$5.00grok-4.7$6.00Claude Sonnet 5.5$10.00Claude Opus 5.5$20.00Claude Opus 4.8$25.00Claude Fable 5.2 (new)$40.00Claude Fable 5.1$50.00
List prices, standard-length prompts, before tax, read Nov 15, 2026. Sources: docs.x.ai and platform.claude.com.

Long-prompt surcharges. Grok models bill the whole request at double the rate once the prompt reaches 200,000 tokens. grok-4.7, for example, goes from $2 / $6 to $4 / $12. Claude Haiku 5.5 rises to $0.50 / $2.50 for prompts over 100,000 tokens. If you send long documents, check which side of those lines your typical request falls on.

Media and voice prices (Grok API, unchanged)

Item Price
grok-imagine-image-2.0 $0.04 per image
grok-imagine-video-1.5 $0.08 per second
grok-imagine-video-1.5-lite $0.02 per second
Grok Voice 3.0, speech to speech $0.08 per minute
Speech to text (REST) $0.10 per hour
Text to speech $15.00 per million characters

Consumer and team plans (unchanged)

The X Subscription Monthly Annual Claude Monthly Annual
Plus $8 $80 Pro $20 $200
Premium (Grok Bot) $30 $300 Max $100 —
Super $100 $1,000 Ultra $200 —
Ultra $200 $2,000

Claude Max and Ultra now include $100 and $200 a month of API credits for building on the Claude Platform. Anthropic’s pricing page labels both as new, and they matter if you are deciding between a plan and a pay-as-you-go API key. All plan prices are before sales tax.

Coming up: dates to put on your calendar

Date Vendor What happens What to do
Nov 30, 2026 Anthropic Opus 4.8 fast mode removed; speed: "fast" on claude-opus-4-8 returns an error Move fast-mode calls to Opus 5.5
Not announced SpaceXSI Grok 4.8 (pre-training done per Musk, Nov 10) No pricing yet; do not budget for it

Members-only analysis

Read the full issue free for a week

The full issue covers 6 sections, 6 linked sources, and a step-by-step “Do this next” list. Your first 7 days are free on any plan, and you can cancel online in two clicks.

$10.00/month or $108.00/year after the free week, plus any sales tax that applies. Or get all 10 topics for $90.00/month.

Quick answers

What is the cheapest Claude model through the API?

As of Nov 15, 2026, Claude Haiku 5.5 is the cheapest at $0.10 per million input tokens and $0.50 per million output tokens for prompts up to 100,000 tokens. Longer prompts are billed at $0.50 and $2.50.

Is Grok cheaper than Claude?

It depends on the tier. Grok 4.7 costs $2 input and $6 output per million tokens. Claude Sonnet 5.5 costs $2 input and $10 output, and Claude Opus 5.5 costs $4 and $20. For a fair comparison, test both models on your own tasks, because token counts differ by model.

Did Grok or Claude subscription prices change this week?

No. Between Nov 9 and Nov 15, 2026, neither The X Subscription nor the Claude Pro, Max, or Ultra plans changed price.

How this issue was made. Prices come from each vendor’s official pricing pages, read on Nov 15, 2026. They are list prices in US dollars, before tax, batch discounts, or negotiated rates. We compared them with the vendors’ dated release notes to find what changed during the week. Created in-house under the direction and review of a real person, with help from AI tools. Agentic Update is not affiliated with any company named here.