DeepSeek
17 listings of 14 models, input listed from $0.036 to $0.90 per million tokens, 3 scored by Artificial Analysis, and the newest listed 17 Oct 2026.
https://openrouter.ai/models·retrieved sha256 · 0bb53a77713f
14 of the 342 priced models, blended 3 in to 1 out, from $0.09 to $1.193. Each bar is one stretch of the catalog on a logarithmic price scale, and the lit part of it is the listings by DeepSeek. The scale is the whole catalog's either way, so the lit part sits where these listings are priced rather than being redrawn to fill the strip.
DeepSeek V4.1 Flash 1017deepseek/deepseek-v4.1-flash-10171.05M$0.036 / $0.50148 / 774ms100%
- Author
- DeepSeek
- Modalities
- text, image → text
- Max output
- 943,718 tokens
- Uptime
- 100% to 95.76%, 51 sellers
- Intelligence
- 28.6
- Coding
- not scored
- Agentic
- not scored
- Epoch index
- 159.4
- Measured over
- 2,266,465 requests, 51 endpoints
- Weights
- on Hugging Face
DeepSeek V4.1 Flash 1017 (batch)deepseek/deepseek-v4.1-flash-1017:batch1.05M$0.036 / $0.15148 / 774ms100%
- Author
- DeepSeek
- Modalities
- text, image → text
- Max output
- 943,718 tokens
- Uptime
- 100% to 95.76%, 51 sellers
- Intelligence
- 28.6
- Coding
- not scored
- Agentic
- not scored
- Epoch index
- 159.4
- Measured over
- 2,266,465 requests, 51 endpoints
- Weights
- on Hugging Face
- Vitna
- not tracked
DeepSeek V4.1 Flashdeepseek/deepseek-v4.1-flash1.05M$0.30 / $1.20150 / 618ms100%
- Author
- DeepSeek
- Modalities
- text, image → text
- Max output
- 943,718 tokens
- Uptime
- 100% to 95.41%, 37 sellers
- Intelligence
- not scored
- Coding
- not scored
- Agentic
- not scored
- Epoch index
- 157.0
- Measured over
- 1,389,152 requests, 37 endpoints
- Weights
- on Hugging Face
DeepSeek V4.1 Flash (batch)deepseek/deepseek-v4.1-flash:batch1.05M$0.112 / $0.336150 / 618ms100%
- Author
- DeepSeek
- Modalities
- text, image → text
- Max output
- 131,072 tokens
- Uptime
- 100% to 95.41%, 37 sellers
- Intelligence
- not scored
- Coding
- not scored
- Agentic
- not scored
- Epoch index
- 157.0
- Measured over
- 1,389,152 requests, 37 endpoints
- Weights
- on Hugging Face
- Vitna
- not tracked
DeepSeek V4 Flash Vision Expdeepseek/deepseek-v4-flash-vision-exp1.05M$0.44 / $1.3259.6 / 1.25s99.88%
- Author
- DeepSeek
- Modalities
- text, image → text
- Max output
- 943,717 tokens
- Uptime
- 99.88% to 0.00%, 4 sellers
- Intelligence
- not scored
- Coding
- 65.0
- Agentic
- 47.5
- Epoch index
- not placed
- Measured over
- 1,988 requests, 4 endpoints
- Weights
- on Hugging Face
- Vitna
- not tracked
DeepSeek V4 Pro 0813deepseek/deepseek-v4-pro-08131.05M$0.111 / $4.4463.4 / 828ms100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 943,718 tokens
- Uptime
- 100% to 99.40%, 21 sellers
- Intelligence
- 23.1
- Coding
- 68.8
- Agentic
- 41.3
- Epoch index
- 156.0
- Measured over
- 104,214 requests, 21 endpoints
- Weights
- on Hugging Face
DeepSeek V4 Flash 0731deepseek/deepseek-v4-flash-07311.05M$0.06 / $0.18115 / 456ms100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 384,000 tokens
- Uptime
- 100% to 72.50%, 26 sellers
- Intelligence
- not scored
- Coding
- 69.1
- Agentic
- 41.0
- Epoch index
- 154.4
- Measured over
- 810,400 requests, 26 endpoints
- Weights
- on Hugging Face
DeepSeek V4 Flash 0731 (batch)deepseek/deepseek-v4-flash-0731:batch1.05M$0.048 / $0.144115 / 456ms100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 384,000 tokens
- Uptime
- 100% to 72.50%, 26 sellers
- Intelligence
- not scored
- Coding
- 69.1
- Agentic
- 41.0
- Epoch index
- 154.4
- Measured over
- 810,400 requests, 26 endpoints
- Weights
- on Hugging Face
- Vitna
- not tracked
DeepSeek V4 Flash 0423deepseek/deepseek-v4-flash1.05M$0.09 / $0.1881 / 2.04s100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 393,216 tokens
- Uptime
- 100% to 90.15%, 13 sellers
- Intelligence
- not scored
- Coding
- 56.2
- Agentic
- 22.2
- Epoch index
- 146.9
- Measured over
- 708,026 requests, 13 endpoints
- Weights
- on Hugging Face
DeepSeek V4 Pro 0423deepseek/deepseek-v4-pro1.05M$0.90 / $1.5059.4 / 1.15s100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 393,216 tokens
- Uptime
- 100% to 81.77%, 13 sellers
- Intelligence
- 20.8
- Coding
- not scored
- Agentic
- not scored
- Epoch index
- 150.2
- Measured over
- 59,755 requests, 13 endpoints
- Weights
- on Hugging Face
DeepSeek V3.2deepseek/deepseek-v3.2164k$0.26 / $0.3826.6 / 1.2s100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 147,456 tokens
- Uptime
- 100% to 99.51%, 4 sellers
- Intelligence
- not scored
- Coding
- 44.2
- Agentic
- not scored
- Epoch index
- 146.3
- Measured over
- 106,812 requests, 4 endpoints
- Weights
- on Hugging Face
DeepSeek V3.2 Expdeepseek/deepseek-v3.2-exp164k$0.27 / $0.4118 / 2.17s100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 147,456 tokens
- Uptime
- 100%, one seller
- Intelligence
- not scored
- Coding
- not scored
- Agentic
- not scored
- Epoch index
- 144.4
- Measured over
- 3,363 requests, 1 endpoint
- Weights
- on Hugging Face
- Vitna
- not tracked
DeepSeek V3.1 Terminusdeepseek/deepseek-v3.1-terminus164k$0.27 / $1.0013 / 1.63s100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 147,456 tokens
- Uptime
- 100%, one seller
- Intelligence
- not scored
- Coding
- 43.5
- Agentic
- 6.8
- Epoch index
- not placed
- Measured over
- 1,950 requests, 1 endpoint
- Weights
- on Hugging Face
- Vitna
- not tracked
DeepSeek V3.1deepseek/deepseek-chat-v3.1164k$0.25 / $0.9550.3 / 409ms100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 147,456 tokens
- Uptime
- 100% to 98.63%, 3 sellers
- Intelligence
- not scored
- Coding
- not scored
- Agentic
- not scored
- Epoch index
- 140.9
- Measured over
- 47,637 requests, 3 endpoints
- Weights
- on Hugging Face
R1 0528deepseek/deepseek-r1-0528164k$0.50 / $2.1529.5 / 1.99s100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 32,768 tokens
- Uptime
- 100% to 99.09%, 3 sellers
- Intelligence
- not scored
- Coding
- not scored
- Agentic
- not scored
- Epoch index
- not placed
- Measured over
- 3,391 requests, 3 endpoints
- Weights
- on Hugging Face
DeepSeek V3 0324deepseek/deepseek-chat-v3-0324164k$0.29 / $1.1420 / 2.09s100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 115,200 tokens
- Uptime
- 100%, one seller
- Intelligence
- not scored
- Coding
- 21.2
- Agentic
- 0.8
- Epoch index
- not placed
- Measured over
- 7,686 requests, 1 endpoint
- Weights
- on Hugging Face
- Vitna
- not tracked
DeepSeek V3deepseek/deepseek-chat164k$0.257 / $1.0350.2 / 1.7s100%
- Author
- DeepSeek
- Modalities
- text → text
- Max output
- 16,000 tokens
- Uptime
- 100% to 82.13%, 3 sellers
- Intelligence
- not scored
- Coding
- not scored
- Agentic
- not scored
- Epoch index
- 131.7
- Measured over
- 18,127 requests, 3 endpoints
- Weights
- on Hugging Face
- Vitna
- priced by Vitna
DeepSeek R1aws-deepseek-r1open weight$1.35 / $5.40AWS
- Made by
- DeepSeek
- Market
- open-weight serving
- OpenRouter
- not listed
- Seller
- AWS $1.35 / $5.40
- Cached input
- not published
- Batch
- not published
- Price since
- Read
- confirmed 22 Nov 14:08 UTC
- Document
- sha256 · f1ac6a175fb2
DeepSeek R1 (Turbo)novita-deepseek-r1-turboopen weight$0.70 / $2.50Novita AI
- Made by
- DeepSeek
- Market
- open-weight serving
- OpenRouter
- no longer lists deepseek/deepseek-r1, first missing
- Seller
- Novita AI $0.70 / $2.50
- Cached input
- not published
- Batch
- not published
- Price since
- Read
- confirmed 22 Nov 14:08 UTC
- Document
- sha256 · eda1835d655d
DeepSeek-OCR 2novita-deepseek-ocr-2open weight$0.03 / $0.03Novita AI
- Made by
- DeepSeek
- Market
- open-weight serving
- OpenRouter
- not listed
- Seller
- Novita AI $0.03 / $0.03
- Cached input
- not published
- Batch
- not published
- Price since
- Read
- confirmed 22 Nov 14:08 UTC
- Document
- sha256 · eda1835d655d
17 listings by DeepSeek, newest listing first. A listing is one OpenRouter id, so a model's batch and free variants are rows of their own, and 17 listings here are 14 models. Prices are OpenRouter's listed rates per million tokens at retrieval, not a provider's own page; the Vitna row inside each listing links the skus whose primary-sourced price Vitna keeps.
Throughput and first-token times are medians of OpenRouter's own routed traffic over a 30-minute window, read . 14 of these 14 models carried a figure; the rest saw no traffic in that window and read as unmeasured, which is not a slow one.
Uptime is OpenRouter's own figure for each seller over the trailing day, read once a day, and the column prints the steadiest seller's reading, taken ; an opened row gives the range down to the lowest, and every seller's own reading is on the model's page.
Beyond the list are the 3 DeepSeek models Vitna prices that OpenRouter does not list, or no longer lists, one row per model however many companies sell it. Each figure is a seller's own list price per million tokens, read from its pricing page and kept as a dated record, and each row leads to that record on the tape.
Source: OpenRouter (openrouter.ai/models), as of 17 Nov 16:29 UTC. sha256 0bb53a77713f over 853,270 bytes.
Catalog: OpenRouter's public model list.
Intelligence, coding and agentic indexes: Artificial Analysis, as published by OpenRouter.
Elo, win rate and rank: Design Arena, as published by OpenRouter.
Source: the Epoch AI Capabilities Index (epoch.ai/eci), as of 20 Nov 2026. 258 models placed over 2917 benchmark results.
Epoch AI Benchmarking Hub data, by Epoch AI, under CC BY 4.0.
Vitna ran none of these evaluations and adjusted none of these figures.
