DeepSeek-R1
Open-weightLast verified: June 2026· DeepSeek DeepSeek-R1 pricing
Open-weight reasoning model. Confirm current pricing with the provider.
- Input / 1M
- $0.55
- Output / 1M
- $2.19
- Context
- 128K tokens
- Provider
- DeepSeek
Pricing verified June 2026. Prices change frequently. Always confirm against the provider’s official pricing page before relying on these figures for budgeting. Official pricing →
When to choose DeepSeek-R1
Open-weight reasoning at $0.55 in and $2.19 out, which makes it the cheapest reasoning option in this table by a wide margin. That price is the reason to try it and the reason to be careful: reasoning routes are where a quality gap shows up as a wrong answer rather than a clumsy one, and a cheap wrong answer on a hard route is not cheap. Run it against your hardest evals before it carries anything load-bearing, and treat the offshore hosted API as a residency decision rather than a footnote.
DeepSeek-R1 sits close enough to DeepSeek-V3 on price that the two are usually decided together. That comparison is written once, on the models table, rather than argued twice from opposite sides.
What DeepSeek-R1 costs in practice
The 50M/10M month is about $49 — a little above Gemini 2.5 Flash’s $40 for a window eight times narrower, and half of o4-mini’s $99. Filling the 128K window costs seven cents of input. The figure to be careful with is the one not shown: reasoning models spend output tokens thinking, so a hard route costs more than this table implies, and the cheaper the model the more tempting it is to skip measuring that. Instrument the token counts on a real route before concluding it is the bargain the headline rate suggests.
| Scenario | Input / mo | Output / mo | Est. cost / mo |
|---|---|---|---|
| Prototype | 2M | 0.5M | $2 |
| Growing product | 50M | 10M | $49 |
| At scale | 500M | 100M | $494 |
Plug in your own numbers with the cost calculator.
Where DeepSeek-R1 fits in a routing setup
Reasoning-shaped work comes to this row and everything else does not; DeepSeek-V3 is the cheaper home for the rest. Upward, o4-mini is the hosted alternative at $99 on the reference month against about $49 — that difference is the price of a first-party API and no residency question. Sideways, Gemini 2.5 Flash does the same month for $40 without being a reasoning model at all, so a route that is not genuinely reasoning-bound is better off there on both counts.
Cheaper alternatives to DeepSeek-R1
The models immediately below DeepSeek-R1 on blended cost, nearest first. A swap only pays if quality holds on the routes you move — see the LLM cost optimization playbook for how to prove that before you move them.
Frequently asked questions
How much does DeepSeek-R1 cost?
DeepSeek-R1 bills $0.55 per million input tokens and $2.19 per million output tokens — $0.96 per million blended at a three-to-one input:output split. A workload of 50M input and 10M output tokens per month works out at about $49. Confirm current pricing with the provider.
What is DeepSeek-R1's context window?
DeepSeek-R1 has a 128K-token context window. One request that fills the whole window costs about $0.07 at the input rate, before the reply is billed.
Where does DeepSeek-R1 sit on price?
At $0.96 per million blended — three input tokens for every output token — DeepSeek-R1 is dearer than 5 of the 19 models tracked here and cheaper than 13. Output bills at 4.0x the input rate, and the nearest cheaper model on this table is Gemini 2.5 Flash at $0.85 blended.
Related concepts: context window, model routing, prompt caching, evals.