vakati-tech.ai

// head_to_head

Claude Haiku 5.5 vs Hy4 flash

Published LiveBench scores across all seven categories, live list pricing, context windows, and the measured cost of a point of capability — for both models, side by side.

OpenRouter + LiveBenchAll comparisonsFull leaderboard

Hy4 flash is the cheaper of the two; neither can be ranked on quality here.

Hy4 flash does not have a published LiveBench run, so this comparison covers price, context and declared capabilities only. A missing score means "not evaluated", not "worse" — the right way to separate these two is an eval on your own workload.

anthropic

Claude Haiku 5.5

Blended / 1M
$0.200
Context
1M
Released
Oct 7, 2026
Overall score
68.9
reasoningtool callingimage inputfile inputprompt caching

tencent

Hy4 flash

Blended / 1M
$0.061
Context
1.0M
Released
Nov 16, 2026
Overall score
Not evaluated
reasoningtool callingimage inputprompt caching

Specs and pricing

MetricClaude Haiku 5.5Hy4 flash
LiveBench overall

Weighted mean of the seven LiveBench category scores, 0–100. Higher is better.

68.9—
Cost per point

Measured benchmark spend divided by overall score — dollars per point of capability.

$0.0126—
Blended price / 1M

3:1 input:output mix, the usual shape of production traffic.

$0.200$0.061win
Input price / 1M$0.100$0.035win
Output price / 1M$0.500$0.140win
Cached input / 1M

Price of an input token served from the prompt cache, where the provider publishes one.

$0.010$0.0015win
Context window1M1.0M
Max output tokens128Kwin66K

Benchmarks by category

An overall score averages away the thing you probably care about. These are the seven LiveBench categories scored separately — Claude Haiku 5.5 on top, Hy4 flash below, both out of 100.

Long-horizon agentic
47.9
—
Coding
77.6
—
Reasoning
82.4
—
Browser
69.5
—
Data analysis
53.1
—
Research
84.3
—
Documents & design
79.3
—

What each one costs to run

Per-token prices are hard to feel. These are monthly list costs for both models across five workload shapes, using each provider's published cached-input rate where there is one.

WorkloadClaude Haiku 5.5Hy4 flash
Support chatbot

1.2K in / 400 out × 200K requests

$57.52/mo$17.23/mo
RAG assistant

8K in / 600 out × 100K requests

$74.00/mo$23.03/mo
Coding agent

40K in / 4K out × 20K requests

$69.60/mo$20.48/mo
Document extraction

20K in / 1.5K out × 50K requests

$133.00/mo$43.86/mo
Bulk classification

500 in / 20 out × 5M requests

$255.00/mo$84.80/mo
Run these two through the cost calculator

Which should you pick?

You are cost-constrained

Hy4 flash

Cheaper on blended list price at $0.061 per million tokens.

Claude Haiku 5.5 vs Hy4 flash FAQ

Which is better, Claude Haiku 5.5 or Hy4 flash?

Hy4 flash is the cheaper of the two; neither can be ranked on quality here. Hy4 flash does not have a published LiveBench run, so this comparison covers price, context and declared capabilities only. A missing score means "not evaluated", not "worse" — the right way to separate these two is an eval on your own workload.

Is Claude Haiku 5.5 cheaper than Hy4 flash?

Hy4 flash is cheaper. On a 3:1 input:output blend, Claude Haiku 5.5 lists at $0.200 per million tokens and Hy4 flash at $0.061 — Hy4 flash is 3.3× cheaper. Input and output are priced separately — Claude Haiku 5.5 charges $0.100 in and $0.500 out, Hy4 flash charges $0.035 and $0.140 — so the model that looks cheaper flips depending on how output-heavy your workload is.

Does Claude Haiku 5.5 or Hy4 flash have a bigger context window?

They are effectively the same — 1M for Claude Haiku 5.5 and 1.0M for Hy4 flash.

Do Claude Haiku 5.5 and Hy4 flash support prompt caching?

Both publish a cached-input rate: $0.010 per million for Claude Haiku 5.5 and $0.0015 for Hy4 flash, against full input rates of $0.100 and $0.035. On a workload with a long stable prefix — a system prompt, a tool schema, a retrieved corpus — that changes the economics more than the headline price does.

Related comparisons

How these numbers are produced

  • Price — provider list price from OpenRouter, refreshed every 15 minutes. “Blended” is a 3:1 input:output mix.
  • Scores — LiveBench release 2026-11-06, using their own category map. Each model shows its strongest published run. A blank means “not evaluated”, never “bad”.
  • Cost per point — the measured dollars LiveBench spent on the run, divided by the score it earned.
  • “Win” — awarded only past a threshold: one full point on a benchmark score, 10% on a price, 25% on a context window. Anything tighter reports as a tie, because effort settings alone move a LiveBench score by more than that.

Published benchmarks rank models on someone else's tasks. Before committing, see LLM & agent evaluation for building an eval on your own.