DeepSeek Models

37 modelsGeneral models from $0.1408/M inputUp to 1M context

Usage

Last 29 days · 2026-10-22 to 2026-11-19

Tokens

1872B

Requests

11M

Models in use

32 of 37

Tokens per day, stacked by model

052.9B106B10-2210-2911-0511-1211-192026-10-22 — 31,479,978,085 tokens deepseek-v4.1-flash-1017: 17,897,177,750 deepseek-v4.1-flash: 9,933,906,495 deepseek-v4-flash: 1,179,589,310 deepseek-v4-flash-0731: 950,664,170 deepseek-v4-pro: 746,200,355 deepseek-v4-pro-0813: 516,407,815 deepseek-v3.2: 138,902,180 6 more models: 104,920,110 deepseek-v4-flash-vision-exp: 12,209,9002026-10-23 — 55,493,545,135 tokens deepseek-v4.1-flash-1017: 26,810,994,350 deepseek-v4.1-flash: 15,883,994,625 deepseek-v4-flash-0731: 9,811,666,215 deepseek-v4-flash: 2,181,944,975 deepseek-v4-pro: 339,686,210 deepseek-v4-pro-0813: 301,541,450 deepseek-v3.2: 101,428,090 deepseek-v4-flash-vision-exp: 47,917,275 6 more models: 14,371,9452026-10-24 — 34,693,073,595 tokens deepseek-v4.1-flash-1017: 24,315,284,605 deepseek-v4.1-flash: 6,668,554,180 deepseek-v4-flash: 1,877,673,700 deepseek-v4-flash-0731: 935,575,145 deepseek-v4-pro-0813: 342,973,545 deepseek-v3.2: 254,562,825 deepseek-v4-pro: 176,110,600 6 more models: 94,538,125 deepseek-v4-flash-vision-exp: 27,800,8702026-10-25 — 21,194,310,620 tokens deepseek-v4.1-flash-1017: 14,897,774,440 deepseek-v4.1-flash: 4,170,969,760 deepseek-v4-flash-0731: 632,847,125 deepseek-v4-pro-0813: 626,072,985 deepseek-v4-flash: 600,604,010 deepseek-v4-pro: 145,559,155 deepseek-v3.2: 100,284,665 deepseek-v4-flash-vision-exp: 16,857,400 6 more models: 3,341,0802026-10-26 — 42,872,003,760 tokens deepseek-v4.1-flash-1017: 30,085,306,115 deepseek-v4.1-flash: 8,419,624,830 deepseek-v4-flash-0731: 1,807,352,320 deepseek-v4-flash: 1,030,279,565 deepseek-v4-pro: 446,755,465 deepseek-v4-pro-0813: 400,931,055 6 more models: 394,713,985 deepseek-v3.2: 261,491,970 deepseek-v4-flash-vision-exp: 25,548,4552026-10-27 — 47,629,807,980 tokens deepseek-v4.1-flash-1017: 31,023,191,010 deepseek-v4.1-flash: 12,248,867,925 deepseek-v4-flash: 1,653,227,465 deepseek-v4-pro-0813: 1,056,737,140 deepseek-v4-flash-0731: 718,745,155 deepseek-v4-pro: 664,754,295 deepseek-v3.2: 122,628,940 deepseek-v4-flash-vision-exp: 120,471,360 6 more models: 21,184,6902026-10-28 — 80,861,631,815 tokens deepseek-v4.1-flash: 47,782,984,610 deepseek-v4.1-flash-1017: 29,499,793,590 deepseek-v4-flash: 1,418,119,535 deepseek-v4-pro: 848,434,525 deepseek-v4-flash-0731: 772,535,790 deepseek-v4-pro-0813: 335,351,240 deepseek-v4-flash-vision-exp: 143,196,110 deepseek-v3.2: 59,300,710 6 more models: 1,915,7052026-10-29 — 77,027,159,520 tokens deepseek-v4.1-flash-1017: 43,241,394,780 deepseek-v4.1-flash: 28,180,868,345 deepseek-v4-flash-0731: 1,680,531,160 deepseek-v4-pro: 1,450,950,460 deepseek-v4-flash: 1,254,900,360 deepseek-v4-pro-0813: 1,028,741,965 deepseek-v3.2: 139,937,070 deepseek-v4-flash-vision-exp: 44,432,200 6 more models: 5,403,1802026-10-30 — 60,246,617,260 tokens deepseek-v4.1-flash-1017: 29,807,202,150 deepseek-v4.1-flash: 26,828,256,215 deepseek-v4-flash-0731: 1,733,136,875 deepseek-v4-flash: 711,765,690 deepseek-v4-pro: 536,068,285 deepseek-v4-pro-0813: 524,626,220 deepseek-v4-flash-vision-exp: 64,858,590 deepseek-v3.2: 21,689,540 6 more models: 19,013,6952026-10-31 — 35,248,191,615 tokens deepseek-v4.1-flash-1017: 20,231,491,280 deepseek-v4.1-flash: 12,937,979,980 deepseek-v4-flash-0731: 773,939,320 deepseek-v4-pro-0813: 559,551,705 deepseek-v4-flash: 507,469,635 deepseek-v4-pro: 206,319,790 6 more models: 15,285,585 deepseek-v4-flash-vision-exp: 12,676,925 deepseek-v3.2: 3,477,3952026-11-01 — 32,508,131,815 tokens deepseek-v4.1-flash: 14,847,846,480 deepseek-v4.1-flash-1017: 14,654,575,635 deepseek-v4-flash-0731: 1,931,850,975 deepseek-v4-flash: 453,803,150 deepseek-v4-pro: 304,619,675 deepseek-v4-pro-0813: 300,987,730 deepseek-v3.2: 7,519,010 6 more models: 5,599,970 deepseek-v4-flash-vision-exp: 1,329,1902026-11-02 — 82,151,017,110 tokens deepseek-v4.1-flash: 47,868,703,045 deepseek-v4.1-flash-1017: 28,570,744,765 deepseek-v4-pro-0813: 2,114,805,430 deepseek-v4-flash-0731: 1,411,757,515 deepseek-v4-flash: 1,290,339,435 deepseek-v4-pro: 864,267,290 deepseek-v3.2: 11,897,170 deepseek-v4-flash-vision-exp: 11,477,575 6 more models: 7,024,8852026-11-03 — 73,818,843,405 tokens deepseek-v4.1-flash-1017: 35,242,137,280 deepseek-v4.1-flash: 34,939,035,855 deepseek-v4-pro: 865,178,560 deepseek-v4-flash: 850,661,680 deepseek-v4-pro-0813: 761,340,390 deepseek-v4-flash-0731: 713,441,725 deepseek-v3.2: 256,108,325 deepseek-v4-flash-vision-exp: 169,515,320 6 more models: 21,424,2702026-11-04 — 66,259,558,785 tokens deepseek-v4.1-flash-1017: 47,602,000,375 deepseek-v4.1-flash: 15,253,863,685 deepseek-v4-flash-0731: 1,391,805,540 deepseek-v4-flash: 891,861,460 deepseek-v4-flash-vision-exp: 367,306,265 deepseek-v4-pro: 346,558,125 deepseek-v3.2: 200,327,765 deepseek-v4-pro-0813: 186,988,910 6 more models: 18,846,6602026-11-05 — 58,936,511,815 tokens deepseek-v4.1-flash-1017: 43,349,105,905 deepseek-v4.1-flash: 11,432,467,480 deepseek-v4-flash: 1,951,299,280 deepseek-v4-pro: 717,040,050 deepseek-v4-flash-0731: 491,736,415 deepseek-v3.2: 375,826,400 deepseek-v4-pro-0813: 355,725,870 deepseek-v4-flash-vision-exp: 249,422,700 6 more models: 13,887,7152026-11-06 — 56,067,744,535 tokens deepseek-v4.1-flash-1017: 35,546,145,975 deepseek-v4.1-flash: 15,951,838,295 deepseek-v4-flash: 1,788,629,295 deepseek-v4-pro-0813: 1,261,285,740 deepseek-v4-pro: 1,022,569,885 deepseek-v4-flash-0731: 283,580,625 deepseek-v4-flash-vision-exp: 147,561,085 deepseek-v3.2: 49,210,795 6 more models: 16,922,8402026-11-07 — 60,175,095,180 tokens deepseek-v4.1-flash: 38,757,703,885 deepseek-v4.1-flash-1017: 18,717,938,030 deepseek-v4-flash: 1,009,310,350 deepseek-v4-pro-0813: 807,735,595 deepseek-v4-flash-0731: 592,414,110 deepseek-v4-pro: 226,185,740 deepseek-v4-flash-vision-exp: 34,980,275 6 more models: 17,634,760 deepseek-v3.2: 11,192,4352026-11-08 — 78,547,724,265 tokens deepseek-v4.1-flash: 45,250,063,890 deepseek-v4.1-flash-1017: 28,215,105,155 deepseek-v4-flash-0731: 2,685,843,690 deepseek-v4-flash: 1,313,468,780 deepseek-v4-pro-0813: 666,703,440 deepseek-v4-pro: 350,591,120 deepseek-v3.2: 29,978,045 deepseek-v4-flash-vision-exp: 26,036,640 6 more models: 9,933,5052026-11-09 — 105,758,698,605 tokens deepseek-v4.1-flash-1017: 50,316,082,570 deepseek-v4.1-flash: 49,064,183,565 deepseek-v4-flash-0731: 2,884,617,750 deepseek-v4-flash: 2,008,336,690 deepseek-v4-pro: 582,889,920 deepseek-v4-pro-0813: 506,998,175 deepseek-v4-flash-vision-exp: 353,633,180 6 more models: 25,320,925 deepseek-v3.2: 16,635,8302026-11-10 — 104,988,443,340 tokens deepseek-v4.1-flash: 52,627,861,430 deepseek-v4.1-flash-1017: 44,129,851,320 deepseek-v4-flash-0731: 2,540,175,745 deepseek-v4-flash: 2,382,078,235 deepseek-v4-pro-0813: 2,240,631,275 deepseek-v4-pro: 690,965,320 deepseek-v4-flash-vision-exp: 289,564,025 deepseek-v3.2: 65,876,935 6 more models: 21,439,0552026-11-11 — 79,209,924,470 tokens deepseek-v4.1-flash-1017: 42,899,439,905 deepseek-v4.1-flash: 24,142,148,890 deepseek-v4-flash-0731: 5,736,699,860 deepseek-v4-pro-0813: 2,346,856,960 deepseek-v4-flash: 2,253,627,790 deepseek-v4-pro: 1,015,824,275 deepseek-v4-flash-vision-exp: 717,413,110 deepseek-v3.2: 90,248,940 6 more models: 7,664,7402026-11-12 — 83,665,744,225 tokens deepseek-v4.1-flash: 41,193,520,485 deepseek-v4.1-flash-1017: 34,939,572,100 deepseek-v4-flash-0731: 3,761,765,295 deepseek-v4-pro: 1,446,282,160 deepseek-v4-flash: 1,324,987,900 deepseek-v4-pro-0813: 702,270,435 deepseek-v3.2: 272,967,935 deepseek-v4-flash-vision-exp: 15,461,670 6 more models: 8,916,2452026-11-13 — 78,792,738,785 tokens deepseek-v4.1-flash-1017: 43,505,758,430 deepseek-v4.1-flash: 23,781,729,375 deepseek-v4-flash-0731: 7,817,630,760 deepseek-v4-flash: 2,731,797,905 deepseek-v4-pro: 699,622,305 deepseek-v4-pro-0813: 160,804,720 deepseek-v3.2: 58,428,430 deepseek-v4-flash-vision-exp: 29,378,880 6 more models: 7,587,9802026-11-14 — 68,615,677,160 tokens deepseek-v4.1-flash-1017: 38,305,402,440 deepseek-v4.1-flash: 18,739,852,860 deepseek-v4-flash-0731: 6,606,504,735 deepseek-v4-flash: 4,055,467,220 deepseek-v4-pro-0813: 740,252,505 deepseek-v4-pro: 105,035,705 6 more models: 37,662,450 deepseek-v3.2: 24,464,265 deepseek-v4-flash-vision-exp: 1,034,9802026-11-15 — 54,773,884,030 tokens deepseek-v4.1-flash-1017: 38,102,618,290 deepseek-v4.1-flash: 12,103,645,830 deepseek-v4-flash: 2,755,492,130 deepseek-v4-flash-0731: 1,260,595,705 deepseek-v4-pro-0813: 233,375,645 deepseek-v4-pro: 154,417,875 deepseek-v4-flash-vision-exp: 76,950,690 6 more models: 73,258,900 deepseek-v3.2: 13,528,9652026-11-16 — 76,287,879,645 tokens deepseek-v4.1-flash-1017: 54,076,111,685 deepseek-v4.1-flash: 16,718,674,255 deepseek-v4-flash: 3,720,419,600 deepseek-v4-flash-0731: 705,811,260 deepseek-v4-pro: 703,034,590 deepseek-v3.2: 146,677,720 deepseek-v4-flash-vision-exp: 119,996,365 deepseek-v4-pro-0813: 89,871,025 6 more models: 7,283,1452026-11-17 — 76,629,557,005 tokens deepseek-v4.1-flash-1017: 56,046,775,575 deepseek-v4.1-flash: 15,295,351,760 deepseek-v4-flash: 3,632,865,885 deepseek-v4-flash-0731: 772,007,135 deepseek-v4-pro-0813: 412,364,315 deepseek-v4-pro: 273,849,680 deepseek-v4-flash-vision-exp: 141,858,445 deepseek-v3.2: 52,315,110 6 more models: 2,169,1002026-11-18 — 82,261,082,215 tokens deepseek-v4.1-flash-1017: 55,826,053,670 deepseek-v4.1-flash: 19,289,866,030 deepseek-v4-flash: 5,149,778,745 deepseek-v4-flash-0731: 1,119,697,755 deepseek-v4-pro-0813: 365,962,290 deepseek-v4-pro: 351,157,125 deepseek-v4-flash-vision-exp: 102,707,700 deepseek-v3.2: 50,280,810 6 more models: 5,578,0902026-11-19 — 65,878,167,550 tokens deepseek-v4.1-flash-1017: 48,598,664,075 deepseek-v4.1-flash: 11,370,431,735 deepseek-v4-flash: 4,628,327,470 deepseek-v4-flash-0731: 451,140,410 deepseek-v4-pro: 451,105,365 deepseek-v4-pro-0813: 179,750,555 deepseek-v4-flash-vision-exp: 135,340,420 deepseek-v3.2: 39,321,250 6 more models: 24,086,270
  • deepseek-v4.1-flash-1017
  • deepseek-v4.1-flash
  • deepseek-v4-flash-0731
  • deepseek-v4-flash
  • deepseek-v4-pro-0813
  • deepseek-v4-pro
  • deepseek-v4-flash-vision-exp
  • deepseek-v3.2
  • 6 more models

Which models that traffic went to

  1. DeepSeek V4.1 Flash 101754.8%1026B
  2. DeepSeek V4.1 Flash36.4%682B
  3. DeepSeek V4 Flash 07313.4%63B
  4. DeepSeek V4 Flash3.0%56.6B
  5. DeepSeek V4 Pro 08131.1%20.1B
  6. DeepSeek V4 Pro0.9%16.7B
  7. DeepSeek V4 Flash Vision Exp0.2%3.5B
  8. DeepSeek V3.20.2%3B
  9. 6 more models0.1%1B

Share of 1872B tokens. 18 models with traffic report no token counts and cannot be ranked here, including deepseek-r1-distill-llama-70b and DeepSeek-V3.1-Think — they are in the request view.

The two views disagree on purpose: a model can take a large share of the calls and a small share of the tokens — many short requests — or the reverse. Which one matters depends on whether your cost is driven by call volume or by prompt length. Measured on AIHubMix over the last 29 days, counting the 37 model IDs listed on this page; traffic routed through upstream-specific IDs that are not in the public catalog is not included.

All 37 DeepSeek Models

Open in model list
DeepSeek models on AIHubMix with input and output modalities, context length, maximum output, price per million tokens including cache read and cache write rates, and measured throughput and latency.
Modalities
deepseek-v4.1-flashTakes text, vision, returns text.1M384K$0.1408$0.5632/M$0.0028/M—96 tok/s1.84 s
deepseek-v4-flashTakes text, vision, returns text.1M393K$0.142$0.284/M$0.0284/M—101 tok/s1.57 s
deepseek-v4-flash-0731Takes text, returns text.1M384K$0.142$0.284/M$0.0284/M—64 tok/s0.66 s
deepseek-v4-flash-vision-expTakes text, vision, returns text.1M393K$0.155$0.62/M$0.0031/M—48 tok/s1.91 s
deepseek-v4.1-flash-1017Takes text, vision, returns text.1M393K$0.155$0.62/M$0.0031/M—119 tok/s0.60 s
deepseek-v4-flash-0731-fastTakes text, returns text.1M384K$0.28$1.4/M$0.07/M—48 tok/s0.66 s
deepseek-v4-pro-0813Takes text, returns text.1M384K$0.6918$2.0754/M$0.0231/M—52 tok/s1.43 s
deepseek-v4-proTakes text, returns text.1M384K$1.69$3.38/M$0.1403/M—75 tok/s1.56 s
deepseek-v3.2-thinkTakes text, returns text.164K64K$0.302$0.453/M$0.0302/M—23 tok/s1.96 s
deepseek-v3.2Takes text, returns text.164K64K$0.58$1.68/M——80 tok/s4.70 s
cc-deepseek-v3Takes , returns text.164K—$0.3$0.3/M——58 tok/s1.61 s
cc-deepseek-v3.1Takes text, returns text.160K—$0.56$1.68/M——59 tok/s0.69 s
deepseek-r1-distill-llama-70bTakes text, returns text.131K—$0.8$1.6/M——101 tok/s1.21 s
DeepSeek-V3.1-ThinkTakes text, returns text.128K32K$0.56$1.68/M——32 tok/s1.29 s
DeepSeek-V3-FastTakes text, returns text.32K—$0.56$2.24/M——150 tok/s1.46 s
DeepSeek-OCRTakes text, vision, returns text.8K—$0.02$0.02/M————
deepseek-ocrTakes text, vision, returns text.8K—$0.02$0.02/M————
deepseek-ai/DeepSeek-R1-Distill-Llama-8BTakes , returns text.——$0.01$0.01/M————
deepseek-ai/DeepSeek-R1-Distill-Qwen-1.5BTakes , returns text.——$0.01$0.01/M————
deepseek-ai/DeepSeek-R1-Distill-Qwen-7BTakes , returns text.——$0.01$0.01/M————
tngtech/DeepSeek-R1T-ChimeraTakes , returns text.——$0.02$0.02/M————
deepseek-ai/DeepSeek-Prover-V2-671BTakes , returns text.——$0.1$0.1/M————
deepseek-ai/DeepSeek-R1-Distill-Qwen-14BTakes , returns text.——$0.1$0.1/M————
deepseek-ai/DeepSeek-Coder-V2-InstructTakes , returns text.——$0.16$0.32/M————
deepseek-ai/deepseek-llm-67b-chatTakes , returns text.——$0.16$0.16/M————
deepseek-ai/DeepSeek-V2-ChatTakes , returns text.——$0.16$0.32/M————
deepseek-ai/DeepSeek-V2.5Takes , returns text.——$0.16$0.32/M————
deepseek-ai/deepseek-vl2Takes , returns text.——$0.16$0.16/M————
deepseek-ai/DeepSeek-R1-Distill-Qwen-32BTakes , returns text.——$0.2$0.2/M————
DeepSeek-v3Takes , returns text.——$0.272$1.088/M——67 tok/s1.17 s
deepseek-v3Takes , returns text.——$0.272$1.088/M——67 tok/s1.17 s
alicloud-deepseek-v3.2Takes , returns text.——$0.274$0.411/M$0.0548/M$0.3425/M35 tok/s1.01 s
azure-deepseek-v3.2Takes , returns text.——$0.58$1.68/M——80 tok/s4.70 s
azure-deepseek-v3.2-specialeTakes , returns text.——$0.58$1.68/M——132 tok/s1.75 s
deepseek-ai/DeepSeek-R1-Distill-Llama-70BTakes , returns text.——$0.6$0.6/M————
deepseek-ai/Janus-Pro-7BTakes , returns text, vision.——$2$2/M————
deepseek-ai/DeepSeek-R1-ZeroTakes , returns text.——$2.2$2.2/M————

Prices are USD per million tokens; cache read and cache write are the rates for prompt-cache hits and for writing a prompt into the cache. Throughput and latency are measured on AIHubMix — the same figures the model detail page shows — not vendor claims. A dash means the catalog does not publish that field for that model, which is not the same as the model not supporting it.

DeepSeek on AIHubMix

Which DeepSeek model should I start with?

deepseek-v4.1-flash at $0.1408/M input — the cheapest entry here that declares tool calling, and it carries a 1M context. Move up to deepseek-ai/DeepSeek-R1-Zero when answer quality matters more than cost, or to deepseek-v4-flash for long-form reasoning.

Which of these models reason before answering?

11 of the 37 models here declare a reasoning phase — they work through the problem before producing an answer, which helps on multi-step problems at the cost of extra output tokens. Use the Reasoning filter above the table to see them. The catalog does not record anything further about how they differ, so this page does not sort them into families.

Why are there several entries for the same model?

Because each row is a route you can call, not a model release. Some IDs name an upstream (azure-, alicloud-, cc-), some are the open-weight repository form (deepseek-ai/…), and some differ only in capitalisation, kept so older integrations keep working.

The catalog does not carry a field saying which of those a given row is, so this page does not sort them into buckets it would have to invent. Every row shows that route’s own price, context and speed — compare those directly, and open a model to see the upstreams that serve it.

How is cached input billed?

The Cache read column is the rate for input tokens served from the prompt cache — for example deepseek-v4-flash-vision-exp bills cache hits at 2% of the input rate and deepseek-v4.1-flash bills cache hits at 2% of the input rate. Cache write is the surcharge for putting a prompt into the cache in the first place, and only a few upstreams bill it separately. A dash in either column means the catalog carries no cache rate for that model, so plan on paying the full input rate.

Do I need a separate DeepSeek account?

No. One AIHubMix key covers every model on this page, and switching between them is a change to the model string — billing, rate limits, and logs stay in one place.

Start calling DeepSeek in one line

One key, one endpoint, 984 models across 49 model authors.