Claude Haiku 5.5 vs Gemini 3.8 Flash

Claude Haiku 5.5 is the cheaper of the two at list prices: $0.10 input and $0.50 output per 1M tokens against $0.75 and $3.75 for Gemini 3.8 Flash. Our feature session costs $0.54 on Haiku and $4.05 on Flash. The Flash price doubles on 1 January 2027, to $1.50 and $7.50.

First model Second model

The Haiku long-prompt rate narrows the gap

Past 100K prompt tokens Haiku 5.5 charges $0.50 and $2.50 per 1M, which is still below the Flash price. Our long agent day sends 150K-token prompts and costs $24.38 on Haiku against $36.56 on Flash, or $6.15 against $9.22 with 90% of prompts cached.

The Flash price change

Google lists $0.75 input and $3.75 output per 1M tokens through 31 December 2026, and $1.50 and $7.50 from 1 January 2027. Cached input is $0.075 today; the note does not give a new cached rate.

Size, inputs and scores

Flash reads up to 1.05M tokens and writes up to 66K; Haiku 5.5 reads 1M and writes up to 128K. Flash also takes video, audio and PDF files.

Haiku 5.5 has published scores: 39.2% on Terminal-Bench 4.0, 46.4% on FrontierCode 1.1 and 57.4% on Humanity's Last Exam with tools. We found none for Gemini 3.8 Flash. Flash was first listed on the API on 2 September 2026; Anthropic released Haiku 5.5 five weeks later, on 7 October.

Side by side

Claude Haiku 5.5 and Gemini 3.8 Flash: specifications and prices
	Claude Haiku 5.5	Gemini 3.8 Flash
Maker	Anthropic	Google
Released	7 October 2026[1]	2 September 2026[2] first API listing
Context window	1M tokens[1]	1.05M tokens[3]
Max output	128K tokens[1]	66K tokens[3]
Input price	$0.10 per 1M tokens[1]	$0.75 per 1M tokens[4]
Output price	$0.50 per 1M tokens[1]	$3.75 per 1M tokens[4]
Cached input	$0.01 per 1M tokens[5]	$0.075 per 1M tokens[4]
Long prompts	Over 100K: $0.50 in, $2.50 out[1] Prompts over 100,000 tokens are billed at the higher rate.	One price
Price change	None announced	Price through 31 December 2026; $1.50 input and $7.50 output per 1M tokens from 1 January 2027.[4]
Open weights	No[1] Offered through the Claude API and cloud platforms; no weights are published.	No[6] Served through the Gemini API; no weights are published.
Reads	text, image[1]	text, image, video, audio, pdf[3]

Benchmarks

None of the benchmarks below has a published score for both models. We only show a score we can cite with its source and date; "Not published" means we found none for that model, not that it scored zero.

Published benchmark scores (higher is better)

Claude Haiku 5.5Gemini 3.8 Flash

Benchmarks: Claude Haiku 5.5 and Gemini 3.8 Flash
Benchmark	Claude Haiku 5.5	Gemini 3.8 Flash
Terminal-Bench 4.0Agentic coding tasks in a real terminal.	39.2%[7]	Not published
FrontierCode 1.1 (Main)Hard software engineering tasks.	46.4%[7]	Not published
Humanity's Last Exam (with tools)Expert-level questions across many fields, tools allowed.	57.4%[7]	Not published

What a coding session costs

Our own estimate from the list prices above, for three sessions we define: a coding agent resends the conversation with every call, so prompts are large and grow. "Cached" assumes 90% of prompt tokens are read from the prompt cache and leaves out cache write fees. A long-prompt price applies when a call's prompt is over its threshold. Work out your own mix with the token cost calculator.

Feature session at list prices, US dollars (lower is cheaper)

Claude Haiku 5.5Gemini 3.8 Flash

Session costs: Claude Haiku 5.5 and Gemini 3.8 Flash
Session	Claude Haiku 5.5	Claude Haiku 5.5 cached	Gemini 3.8 Flash	Gemini 3.8 Flash cached
Quick fix10 calls of 25K prompt and 1.5K output tokens	$0.03	$0.01	$0.24	$0.09
Feature session60 calls of 80K prompt and 2K output tokens	$0.54	$0.15	$4.05	$1.13
Long agent day300 calls of 150K prompt and 2.5K output tokens	$24.38	$6.15	$36.56	$9.22

Value for money and release gap

Claude Haiku 5.5 is dated 5 weeks later (35 days; Gemini 3.8 Flash: first API listing). Price per point divides our feature-session cost by a published score: lower means more score for the money. It is a rough lens, not a ranking.

Value for money: Claude Haiku 5.5 and Gemini 3.8 Flash
Measure	Claude Haiku 5.5	Gemini 3.8 Flash
Input tokens per $1	10M	1.33M
Filling the whole context window once	$0.50	$0.79

Questions

Is Claude Haiku 5.5 cheaper than Gemini 3.8 Flash?

Yes, at list prices: $0.10 against $0.75 per 1M input tokens and $0.50 against $3.75 per 1M output, for prompts up to 100K tokens. Above that Haiku charges $0.50 and $2.50, still below Flash.

Will Gemini 3.8 Flash get more expensive?

Its listed price runs through 31 December 2026. From 1 January 2027 it is $1.50 per 1M input tokens and $7.50 per 1M output.

Which has published benchmark scores?

Claude Haiku 5.5: 39.2% on Terminal-Bench 4.0, 46.4% on FrontierCode 1.1 and 57.4% on Humanity's Last Exam with tools. We found none for Gemini 3.8 Flash.

Most compared

Any other two models: pick them in the lists at the top of the page.

More tools

Sources

platform.claude.com: docs/en/models/haiku-5-5/overview, read 22 October 2026
openrouter.ai: google/gemini-3.8-flash, read 22 October 2026
ai.google.dev: gemini-api/docs/models/gemini-3.8-flash, read 22 October 2026
ai.google.dev: gemini-api/docs/pricing, read 22 October 2026
platform.claude.com: docs/en/about-claude/pricing, read 22 October 2026
ai.google.dev: gemini-api/docs/models, read 22 October 2026
anthropic.com: claude-haiku-5-5, read 22 October 2026

Numbers from the makers' own pages, checked by hand, last read 22 October 2026. List prices in US dollars before tax. Spotted a mistake? office@shuriken.sh.

Made by Shuriken, the harness for coding agents. Its hot swapping moves your chats to another of your accounts or API keys when one nears its limit. Use only accounts you own and are entitled to use.

Try Shuriken on your own computer

Add your Claude, Codex and Gemini accounts once. Shuriken switches between them so your agents keep working.

Download Pricing

Coming soon