Claude Haiku 5.5 vs Gemini 4 Carbon

vs

什么时候选哪个

claude-haiku-5-5 和 gemini-4-carbon 都支持在约一百万 token 的上下文(分别为 1000000 和 1048576)中读取文本和图像,但两者的定价相差甚远:前者的输入和输出分别为 $0.1 和 $0.5,后者则为 $2 和 $10,两端均相差 20 倍;前者的缓存读取费率为 $0.01,而后者为 $0.1。如果要处理大批量的文本和图像任务、需要长达 128000 token 的回复,或者希望能够关闭思考功能,请选择 claude-haiku-5-5。当输入包含音频或视频时,请选择 gemini-4-carbon,因为 claude-haiku-5-5 不支持这两种输入。

Benchmark 成绩

Claude Haiku 5.5:供应商没有公布 benchmark 成绩。

高于同类均值无更高分Gemini 4 Carbon12 / 1210 / 12
Claude Haiku 5.5 Gemini 4 Carbon 其他参测模型 同类均值 ★ 没有模型得分更高
FrontierSWE 2 Proximal leaderboard, Carbon report
N/A
61.2%
OSWorld 2.1 82 offline tasks, partial score. Gemini: best of 3. Peers: published evaluations
N/A
81.5%
AutomationBench (Zapier) Zapier private set
N/A
没有模型得分更高 53.9%
GraphWalks BFS (F1) 256k to 1M tokens, 200 tasks, Google evaluation
N/A
没有模型得分更高 96.2%
Blueprint-Bench 2 Similarity score, Andon Labs leaderboard
N/A
没有模型得分更高 58.6%
Vending-Bench 2 (USD) Andon Labs leaderboard, Carbon report · 5422-16484
N/A
没有模型得分更高 16484
Chartography No tools, Carbon report
N/A
没有模型得分更高 72.5%

供应商公布: Google

价格

Claude Haiku 5.5 Gemini 4 Carbon Δ
输入 / 1M token $0.1 $2 0.05×
输出 / 1M token $0.5 $10 0.05×
缓存读取 / 1M token $0.01 $0.1 0.1×
缓存写入 1.25x (5m) / 2x (1h) - -

价格取自构建时的实时目录,最新价格见各模型页面。 主表显示基础档。提示词长度可能改变整次请求的费率,各模型页有完整价目。

Claude Haiku 5.5 · 按输入长度分档的价格

整次请求按输入长度所在档位计费。价格单位为美元/百万 token。

输入 token 数输入输出
0 - 100,000 $0.1 $0.5
100,001 - 1,000,000 $0.5 $2.5

两个模型所处的位置:全部 86 个按同一单位计费的聊天模型的每 1M token 输入价分布(对数刻度)

能力

Claude Haiku 5.5 Gemini 4 Carbon
工具调用 是 是
思考控制 可配置 始终开启
结构化输出 是 是
提示词缓存 显式(由你标记前缀) 隐式(自动)
缓存有效期 5m default, 1h option 未公布
最小缓存前缀 512 个 token 4,096 个 token

规格

Claude Haiku 5.5 Gemini 4 Carbon
输入模态 文本 图像 文本 图像 音频 视频
输出模态 文本 文本
发布日期 2026-10-07 2026-11-10
知识截止日期 2026-06 2026-03
上下文窗口 1,000,000 1,048,576
最大输出 128,000 65,536
思考参数 output_config.effort thinking_level
可选值
output_config.effort
  • low
  • medium
  • high
  • xhigh
  • max
thinking_level
  • low
  • medium
  • high

"minimal" and thinking_budget return 400

默认值 medium medium

规格照录自各供应商的文档;供应商没有公布的项目,对应的行直接省略,不做推断。 完整来源: Claude Haiku 5.5 · Gemini 4 Carbon

同一条提示词,两个模型,经网关实测

提示词 Explain to a CFO, in exactly three sentences, why our LLM bill doubled after we enabled reasoning. 检查 恰好 3 句

Claude Haiku 5.5 通过 · 3 sentences

Enabling reasoning makes the model generate a long internal chain of thought before its answer, and those thinking tokens are billed as output tokens, which cost several times more than input tokens. The model often produces many times more output per request than it did without reasoning, so the per-call cost rises even when the visible answer looks about the same length. Because that extra spend scales with every request, the bill roughly doubled, and we can control it by setting reasoning budgets, using reasoning only for tasks that need it, or routing simpler requests to a non-reasoning model.

输出 332 tok (+159 思考) 延迟 2.9 s

Gemini 4 Carbon 通过 · 3 sentences

Enabling reasoning instructs the AI to generate a lengthy, behind-the-scenes "thought process" to verify its logic before delivering a final answer. Because our vendor bills us based on total output volume, we are charged for all of that invisible scratchpad work in addition to the visible response, effectively doubling the compute consumed per query. Moving forward, we can control this cost by restricting reasoning strictly to complex, high-stakes workflows where higher accuracy justifies the premium, while routing routine tasks back to standard, lower-cost models.

输出 908 tok (+799 思考) 延迟 18.0 s

指令遵循(恰好三句,数得出来)、受众适配(对 CFO 说话的口吻),以及下方 token 计数暴露出的隐藏思考计费差额。

提示词 This function is supposed to return the longest run of consecutive equal items, but callers report it is off by one on single-element inputs and misses runs that end at the last index. Fix it and explain the bug in one paragraph. 检查 修复通过测试

Claude Haiku 5.5 通过 · 8/8 cases

```python def longest_run(xs): if not xs: return 0 best = cur = 1 for i in range(1, len(xs)): if xs[i] == xs[i - 1]: cur += 1 else: cur = 1 best = max(best, cur) return best ``` The bug is that `cur` counts adjacent equal *pairs*, not run *elements*. A run of k equal items produces k−1 matching pairs, so the function returns one less than the true length, and a single-element input produces no pairs at all, so it returns 0 instead of 1. Resetting `cur` to 0 on a mismatch compounds this by starting every run one step behind. The fix starts both `cur` and `best` at 1 (a non-empty list always has a run of at least one element), resets `cur` to 1 on a mismatch, and updates `best` after every element rather than only inside the equal-branch. The empty-list guard keeps the function from returning 1 for `[]`. […]

输出 1012 tok (+591 思考) 延迟 6.0 s

Gemini 4 Carbon 通过 · 8/8 cases

```python def longest_run(xs): if not xs: return 0 best = 1 cur = 1 for i in range(1, len(xs)): if xs[i] == xs[i - 1]: cur += 1 best = max(best, cur) else: cur = 1 return best ``` The original function counts the number of matching adjacent *transitions* rather than the number of *elements* in a run because `best` and `cur` are initialized and reset to `0` instead of `1`. Because a single item on its own already forms a run of length `1`, starting at `0` causes every non-empty result to be short by one: single-element lists (where the loop never executes) and lists with no duplicates return `0` instead of `1`, and any run of $k$ equal items—including one ending at the last index—only increments `cur` $k - 1$ times, making it appear as though the final element of the run was missed. […]

输出 1411 tok (+1131 思考) 延迟 32.7 s

修复是否真的正确(能运行)、解释的信息密度,以及在一个范围明确的任务上的 token 效率。

提示词 Extract a JSON object with fields {company, quarter, revenue_usd_m, growth_pct, guidance} from this text. Use null for anything not stated; add no other fields. 检查 合法 JSON,严格符合 schema

Claude Haiku 5.5 通过 · 5/5 fields, guidance "Formal guidance withheld pending the Q4 "

```json { "company": "Meridian Robotics", "quarter": "September quarter", "revenue_usd_m": null, "growth_pct": 11, "guidance": "Formal guidance withheld pending the Q4 close" } ``` Notes: The text gives revenue as 412 but doesn't state the currency, so I left `revenue_usd_m` as null. Growth is "eleven and change," so 11 is an approximation.

输出 725 tok (+572 思考) 延迟 4.6 s

Gemini 4 Carbon 通过 · 5/5 fields, guidance null

```json { "company": "Meridian Robotics", "quarter": "September quarter", "revenue_usd_m": null, "growth_pct": null, "guidance": null } ```

输出 3338 tok (+3281 思考) 延迟 61.6 s

是否严守 schema(不臆造字段)、能否顶住幻觉压力(原文明说暂不给出 guidance),以及结构化输出路径的差异。

提示词 Write a 120-word product announcement for an API gateway feature that caches prompts across providers. Forbidden words: "seamless", "unlock", "game-changer", "revolutionize", "empower". Exactly one sentence must be a question. 检查 120 词,0 个禁用词

Claude Haiku 5.5 通过 · 120 words, 0 banned, 1 question

**Introducing Prompt Cache Sync, the newest feature in GatewayOne.** Teams routing requests across OpenAI, Anthropic, and Google have often paid to process the same long system prompts again and again. Now the gateway stores those prompts once and reuses them wherever they appear, regardless of provider. Repeated context is matched by content hash, so cached prefixes return faster and cost less, with no changes to your application code. How much could this save your team? Enable Prompt Cache Sync from the dashboard, and your next billing report will show the difference. Cache entries follow your existing retention settings and can be purged per project at any time. Availability begins this week for all Scale and Enterprise plans worldwide. Start today.

输出 2164 tok (+1929 思考) 延迟 9.9 s

Gemini 4 Carbon 通过 · 120 words, 0 banned, 1 question

Are your engineering teams paying multiple AI vendors to generate identical responses? Today, we are introducing Universal Prompt Caching for our flagship API Gateway. When applications route traffic across providers like OpenAI, Anthropic, and Mistral, duplicate queries drain budgets and degrade user experiences. Our new unified caching layer intercepts incoming prompts directly at the network edge. If an identical request was previously fulfilled by any connected model, the gateway delivers the stored response in milliseconds, completely bypassing external API calls. This practical upgrade cuts monthly token expenses substantially while accelerating overall application speed. Best of all, deployment requires zero complex code changes. Simply activate this capability inside your gateway configuration dashboard today to start reducing latency and infrastructure costs immediately.

输出 2620 tok (+2480 思考) 延迟 31.1 s

是否守住约束(字数预算、禁用词表、只能有一个问句)、文风特征,以及长度控制。

改一行代码就能在两个模型之间切换

下方每个标签页里都有两个模型 ID,高亮的那两行是唯一要改的地方。端点不变,API key 不变,请求结构也不变。

from openai import OpenAI

client = OpenAI(
    base_url="https://synthorai.io/v1",
    api_key="sk-syn-...",
)

resp = client.chat.completions.create(
    model="claude-haiku-5-5",
    # model="gemini-4-carbon",  # 取消注释此行,注释上一行
    messages=[{"role": "user", "content": "Summarize this diff"}],
    reasoning_effort="medium",
)
print(resp.choices[0].message.content)

获取 API key →

常见问题

Claude Haiku 5.5 和 Gemini 4 Carbon 哪个更便宜?

主表显示基础档。提示词长度可能改变整次请求的费率,各模型页有完整价目。

不用分别集成两次,就能对 Claude Haiku 5.5 和 Gemini 4 Carbon 做 A/B 测试吗?

可以。两个模型走同一个 OpenAI 兼容端点,用同一个 API key,切换时只要改一行里的模型名,所以可以给两个模型各分一部分流量,直接对比账单。

Claude Haiku 5.5 和 Gemini 4 Carbon 支持提示词缓存吗?

支持。两个模型的缓存读取价都低于各自的输入价,所以前缀能反复命中缓存的负载,实际成本会比按官网价估算的低。具体的缓存读取价见上方价格表。

相关对比

我们的实测研究