Claude 3.5 Haiku
Anthropic · text, image → text
Claude 3.5 Haiku is the next generation of Claude's fastest model.
Input$1.1 /M
Output$5.5 /M
← ModelsClaude 3.5 HaikuClaude Opus 4.8
Claude 3.5 Haiku vs Claude Opus 4.8
Compare Claude 3.5 Haiku from Anthropic and Claude Opus 4.8 from Anthropic on key metrics including benchmarks, price, context length, and other model features. Access both models and hundreds of others through the AIHubMix API.
Claude 3.5 HaikuClaude Opus 4.8+ Add model · 2/4
Claude Opus 4.8
Anthropic · text, image, pdf → text
Claude Opus 4.8 is Anthropic’s newest and most powerful publicly available model. It is suitable for the most complex tasks. It is Anthropic’s strongest model for complex reasoning, long-horizon agent programming, and highly autonomous work.claude-opus-4-8 does not display thought content by default; you need to set the additional parameter "display": "summarized" to enable it
Input$5 /M
Output$25 /M
Cache read$0.5 /M
Pricing & Specifications
Prices are per million tokens. Time to First Token and throughput are rolling averages measured on AIHubMix.
Claude 3.5 Haiku
Claude Opus 4.8
Input /M
$1.1
$5
Output /M
$5.5
$25
Cache read /M
-
$0.5
Context length
200,000
1,000,000
Max output
8,192
128,000
Time to First Token
1.0 s
1.1 s
Throughput
62.4 tok/s
82.6 tok/s
Modalities
textimage
textimagepdf
Supported Parameters
toolsfunction callingstructured outputs
thinkingtoolsfunction callingstructured outputs
API Formats
chat_completions · claude_api
Released
-
May 28, 2026
Promotional prices show the discounted rate; see each model page for promotion windows.
Activity Past 30 Days
Daily traffic served through AIHubMix — how demand for each model is trending.
claude-3-5-haikuclaude-opus-4-8
Tokens / day
-
Requests / day
-
Performance Past 3 Days
Measured on real AIHubMix traffic, hourly buckets. Gaps mean no traffic in that hour.
claude-3-5-haikuclaude-opus-4-8
Throughput (tok/s)
-
TTFT (s)
-
Uptime (%)
-
LMArena Benchmarks
LMArena ratings by capability (Bradley-Terry, commonly called Elo). Higher is better.
Text
claude-3-5-haikuclaude-opus-4-8
1260134014201500
Overall
13231473
Coding
13851527
Math
12861474
Hard prompts
13441503
Instruction following
13141477
Multi-turn
13261495
Creative writing
13041465
Longer query
13431499
Chinese
13151523
English
13361480
Vision
claude-3-5-haikuclaude-opus-4-8
11001160122012801340
Overall
11261279
OCR
11261294
Diagram
11311309
Homework
11421320
Source: LMArena (arena.ai) leaderboard, imported by AIHubMix. Models without published ratings are omitted per chart.
Cost calculator
Estimate your monthly bill for the same workload on each model.
Claude 3.5 Haiku
$149 /mo
Claude Opus 4.8
$675 /mo
Monthly = daily × 30. Discounted rates applied where a promotion is active.
FAQ
Which is cheaper: Claude 3.5 Haiku, Claude Opus 4.8?
Claude 3.5 Haiku: $5.5/M output tokens; Claude Opus 4.8: $25/M. Use the cost calculator above to estimate your own workload.
How do their coding arena scores compare?
Claude Opus 4.8: 1527; Claude 3.5 Haiku: 1385 (LMArena coding leaderboard).
Which responds faster?
Claude 3.5 Haiku: 1.0s time to first token measured on AIHubMix; see the live performance charts above for how each model behaves across the day.
How large is each context window?
Claude 3.5 Haiku accepts 200,000 and Claude Opus 4.8 accepts 1,000,000 input tokens. Maximum output per request is 8,192 tokens on Claude 3.5 Haiku and 128,000 tokens on Claude Opus 4.8.
Which one generates tokens faster?
Claude Opus 4.8 at 82.6 tok/s and Claude 3.5 Haiku at 62.4 tok/s, measured as output throughput on AIHubMix — a separate metric from time to first token.
What inputs and capabilities does each model support?
Claude 3.5 Haiku accepts text and image input and supports tool calling, function calling and structured outputs; Claude Opus 4.8 accepts text, image and PDF input and supports thinking, tool calling, function calling and structured outputs.
Can I call Claude 3.5 Haiku and Claude Opus 4.8 with the same API key?
Yes. AIHubMix serves every model on this page behind one OpenAI-compatible endpoint, so switching between them is a one-line change to the model field — no second account, key or SDK.
Popular comparisons
Related model match-ups readers also look at.
