新帳號 免費註冊,送 10 次呼叫,最高 $1,免綁卡。

DeepSeek V4 Pro (0813) vs Gemini 3.6 Flash

vs

什麼情況選哪一個

若要進行高流量的純文字推理工作,請選擇 deepseek-v4-pro-0813:它將推理做為一項明確的功能,每百萬 tokens 輸入 $1.32 且輸出 $3.96(輸出大約比 $7.5 的 gemini-3.6-flash 便宜 1.9x),並允許高達 393216 輸出 tokens,大約是 Google 模型 65536 上限的 6x。當輸入不只有文字時請選擇 gemini-3.6-flash——它除了文字之外還接受圖片、影片與音訊,並加入視覺能力,音訊輸入按每百萬音訊 tokens $5 計費。在上下文方面 1000000 對比 1048576 tokens 實際上並無明顯差異;請注意,在 gemini-3.6-flash 上無法停用思考(thinking)。

Benchmark 成績

高於平均沒有模型更高DeepSeek V4 Pro (0813)13 / 211 / 21Gemini 3.6 Flash6 / 181 / 18
DeepSeek V4 Pro (0813) Gemini 3.6 Flash 其他有成績的模型 其他模型平均 ★ 沒有模型分數更高
DeepSWE 1.1
62.7%
48.6%
BioMysteryBench hard
N/A
41.2%
OSWorld 2.0
N/A
33.8%
Cybergym
83.3%
N/A
GDPval-AA v2 Elo · 1508-1769 據 Z.ai 公布 · 2026-09-04
1590
N/A
Harvey Lab-AA
N/A
85.1%
GPQA Diamond
92.4%
N/A
Agents' Last Exam
25.7%
N/A
LVBench
N/A
84.2%

供應商公布: Alibaba (Qwen) Anthropic ByteDance DeepSeek Google MiniMax Moonshot OpenAI Tencent Z.ai

定價

DeepSeek V4 Pro (0813) Gemini 3.6 Flash Δ
輸入 / 1M tokens $1.32 $1.5 0.88×
輸出 / 1M tokens $3.96 $7.5 0.53×
快取讀取 / 1M tokens $0.132 $0.15 0.88×
快取寫入 不額外計費 - -

費率取自網站建置時的即時目錄;各模型頁面都列有最新的價目。

兩者的相對位置:每 1M tokens 的輸入價格,涵蓋同一計費單位下全部 76 個聊天模型(對數尺度)

功能

DeepSeek V4 Pro (0813) Gemini 3.6 Flash
工具使用 是 是
思考控制 一律開啟 一律開啟
結構化輸出 是 是
提示詞快取 隱式(自動) 隱式 + 顯式
快取存活時間 no fixed TTL (evicted when unused) 未公布
最小快取前綴 未公布 4096 個 token

規格

DeepSeek V4 Pro (0813) Gemini 3.6 Flash
輸入模態 文字 文字 圖像 音訊 影片
輸出模態 文字 文字
發布日期 2026-08-13 2026-07-21
知識截止日期 - 2026-03
上下文視窗 1M 1M
最大輸出 393K 66K
思考參數 reasoning_effort thinkingLevel
可接受的值
reasoning_effort
  • the model card documents low
  • high
  • max
  • minimal
  • low
  • medium
  • high
預設值 - medium

規格摘錄自各供應商的文件;供應商沒有公布的項目就直接略過,不自行推測。 完整來源: DeepSeek V4 Pro (0813) · Gemini 3.6 Flash

同一段提示詞,兩個模型,經閘道實測

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

DeepSeek V4 Pro (0813) 通過 · 3 sentences

Enabling reasoning causes the model to generate a hidden chain-of-thought with many additional tokens before producing the final answer, which sharply increases compute consumption per request. Those extra reasoning tokens are billed at the same or higher rates, so total usage doubled even though the volume of final outputs stayed similar. We can control this by adjusting reasoning effort, setting token limits, or routing only complex requests through reasoning mode.

輸出 308 tok (+226 思考) 延遲 5.1 s

Gemini 3.6 Flash 通過 · 3 sentences

Enabling reasoning causes the model to generate thousands of internal "thinking" tokens behind the scenes to process logic before delivering the final response. Because our vendor bills us for every token generated—including these intermediate processing steps—our billable output volume per query surged significantly. Since output tokens carry a much higher unit cost than input tokens, this jump in processing volume directly doubled our overall spend.

輸出 893 tok (+816 思考) 延遲 4.7 s

是否照指令做(剛好三句,數得出來)、語氣是否貼合對象(寫給 CFO 看),以及下方 token 計量顯示出的隱藏思考計費落差。

提示詞 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. 檢查 修正後通過測試

DeepSeek V4 Pro (0813) 通過 · 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 else: cur = 1 best = max(best, cur) return best ``` The bug is that `cur` was initialized to `0` and only counted adjacent matching pairs, so a run of length `n` produced `n - 1` (or `0` for a single element) instead of `n`. The code also reset to `0` on a change, when the new run should start at length `1`. Initializing `cur` and `best` to `1` fixes both the single-element case and runs ending at the last index.

輸出 1315 tok (+1130 思考) 延遲 16.9 s

Gemini 3.6 Flash 通過 · 8/8 cases

Here is the fixed function along with an explanation of the bugs: ```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 else: cur = 1 best = max(best, cur) return best ``` ### Explanation The original function initialized `cur` and `best` to `0` and incremented `cur` on matching adjacent items, which meant it was counting identical adjacent *pairs* rather than the total number of *elements* in a run. This caused single-element inputs to skip the loop and return `0` instead of `1`, and undercounted all run lengths by 1 (for example, a run of two items at the end of the list resulted in `cur = 1`). […]

輸出 2034 tok (+1768 思考) 延遲 10.0 s

修正是否真的正確(跑得起來)、說明的資訊密度,以及在範圍明確的任務上的 token 效率。

提示詞 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. 檢查 有效的 JSON,完全符合 schema

DeepSeek V4 Pro (0813) 通過 · 5/5 fields, guidance null

{"company":"Meridian Robotics","quarter":"September quarter","revenue_usd_m":412,"growth_pct":11.0,"guidance":null}

輸出 2155 tok (+2121 思考) 延遲 26.0 s

Gemini 3.6 Flash 通過 · 5/5 fields, guidance "withheld"

```json { "company": "Meridian Robotics", "quarter": "September quarter", "revenue_usd_m": 412, "growth_pct": null, "guidance": "withheld" } ```

輸出 2843 tok (+2783 思考) 延遲 13.1 s

是否遵守 schema(不自創欄位)、面對幻覺誘因的表現(題目明確不給 guidance),以及結構化輸出路徑的差異。

提示詞 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. 檢查 120 詞,0 個禁用詞

DeepSeek V4 Pro (0813) 通過 · 121 words, 0 banned, 1 question

Introducing cross-provider prompt caching in our API gateway. Send a prompt once, and the gateway stores its computed prefix across Anthropic, OpenAI, and other LLM providers. Subsequent requests with the same prompt hit the cache, cutting latency and token costs while keeping outputs consistent across routing decisions and provider failovers. Teams can route identical prompts between providers without reprocessing shared context or lengthy system instructions. How much could you save on repeated prompt prefixes? The cache respects provider-specific key formats, handles TTLs automatically, and works with streaming and batch requests. Enable it with one configuration flag—no changes to your application code. Available today on all plans. Monitor cache hit rates, token savings, and provider-specific performance metrics in the live dashboard.

輸出 2845 tok (+2694 思考) 延遲 25.5 s

Gemini 3.6 Flash 未通過 · 60 words, 0 banned, 0 questions

72: reducing 73: your 74: monthly 75: token 76: spend. S6 (21): 77: You 78: can 79: easily 80: set 81: custom 82: expiration 83: rules, 84: configure 85: TTL 86: settings, 87: and 88: manage 89: cache 90: invalidation 91: across 92: all 93: vendors 94: from 95: one 96: centralized 97: dashboard. S7 (23): 98: Start 99: optimizing […]

輸出 4092 tok (+3929 思考) 延遲 16.3 s

是否遵守限制(字數上限、禁用詞清單、只能有一句問句)、文字風格的特徵,以及長度控制。

改一行程式碼就能在兩者之間切換

下面每個頁籤都列了這兩個模型 ID,要改的只有醒目標示的那兩行。端點、金鑰和請求格式都不變。

from openai import OpenAI

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

resp = client.chat.completions.create(
    model="deepseek-v4-pro-0813",
    # model="gemini-3.6-flash",  # 取消這一行的註解,並把上一行註解掉
    messages=[{"role": "user", "content": "Summarize this diff"}],
    reasoning_effort="medium",
)
print(resp.choices[0].message.content)

取得 API 金鑰 →

常見問題

DeepSeek V4 Pro (0813) 和 Gemini 3.6 Flash 哪個比較便宜?

以「輸入 / 1M tokens」來看,DeepSeek V4 Pro (0813) 比較便宜($1.32 對 $1.5,相差 1.1×)。其他項目的結果可能相反,完整價目請看上表;實際成本要看你的用量組合。

可以只串接一次,就對 DeepSeek V4 Pro (0813) 和 Gemini 3.6 Flash 做 A/B 測試嗎?

可以。兩個模型都走同一個 OpenAI 相容端點,用的也是同一把 API 金鑰,切換時只要改一行裡的模型名稱字串。你可以把一部分流量分別導到兩邊,再直接比較帳單。

DeepSeek V4 Pro (0813) 與 Gemini 3.6 Flash 支援提示詞快取嗎?

支援。兩者的快取讀取費率都低於輸入費率,所以前綴已經進快取的工作負載,實際成本會比官網價算出來的低。確切的快取讀取價格請見上方定價表。

相關比較