Claude Fable 5 vs Claude Haiku 4.5

Benchmarks, API pricing and specs, head to head. Data updated 2026-06-10.

Claude Fable 5

Anthropic · Jun 2026

The first GA model of Anthropic's Mythos class — a tier above Opus — and the current SWE-bench Verified leader, at $10/$50 per 1M tokens.

Claude Haiku 4.5

Anthropic · Oct 2025

Anthropic's fast tier: Sonnet-4-class coding performance at $1/$5 and more than twice the speed.

The verdict

Claude Fable 5 wins 1 of the 1 benchmarks these models share, against 0 for Claude Haiku 4.5. On price the gap is dramatic: Claude Haiku 4.5 works out roughly 10x cheaper per blended million tokens. Claude Fable 5 also takes 1M of context versus 200K for Claude Haiku 4.5.

Benchmark head-to-head 10

SWE-bench Verified
~95%73.3%
Claude Fable 5Claude Haiku 4.5

Specs & pricing

Claude Fable 5Claude Haiku 4.5
modhub Index
Input price / 1M$10$1
Output price / 1M$50$5
Context window1M200K
Max output128K64K
Open weightsnono
Reasoning modelyesyes
Multimodal inputtext, imagetext, image
Knowledge cutoffJan 2026Feb 2025
ReleasedJun 2026Oct 2025
Example monthly cost*$175.00$17.50

* 10M input + 1.5M output tokens per month at list prices, no caching. Green = better value on that row.

Frequently asked questions

Which is better, Claude Fable 5 or Claude Haiku 4.5?
Claude Fable 5 wins 1 of the 1 benchmarks these models share, against 0 for Claude Haiku 4.5. On price the gap is dramatic: Claude Haiku 4.5 works out roughly 10x cheaper per blended million tokens. Claude Fable 5 also takes 1M of context versus 200K for Claude Haiku 4.5.
Which is cheaper, Claude Fable 5 or Claude Haiku 4.5?
Claude Fable 5 costs $10/$50 per million input/output tokens, while Claude Haiku 4.5 costs $1/$5. For a typical workload of 10M input and 1.5M output tokens per month, that's $175.00 versus $17.50.
Which model is better for coding, Claude Fable 5 or Claude Haiku 4.5?
On SWE-bench Verified — the standard agentic-coding benchmark — Claude Fable 5 scores ~95% versus 73.3% for Claude Haiku 4.5, making Claude Fable 5 the stronger pick for coding agents.

More comparisons