Affordable reasoning champions: DeepSeek R1 ($0.55/$2.19) vs OpenAI o3-mini ($1.10/$4.40).
deepseek-reasoner · deepseek
o3-mini · openai
Benchmark
| Workload Scenario | DeepSeek R1 (Reasoner) | o3-mini | Price Delta |
|---|---|---|---|
| 1M input tokens (raw text) | $0.55 | $1.10 | −$0.55 |
| 1M output tokens (generation) | $2.19 | $4.40 | −$2.21 |
| 1M tokens · 70% input / 30% output mix | $1.042 | $2.09 | −$1.048 |
| Standard chat turn (4K in / 800 out, 50% cached) | $0.003132 | $0.00682 | −$0.003688 |
| Monthly scale (10K requests / day) | $939.60 | $2,046.00 | −$1,106.40 |
Negative difference = DeepSeek R1 (Reasoner) is cheaper. Positive = o3-mini is cheaper.
Scaling Curve
Total cost of a token volume at 70/30 input/output split (uncached). Log-log scale.
Verdict
DeepSeek R1 (Reasoner) is cheaper on both input and output rates, so it costs less at every input/output mix. Price alone still isn't the whole decision: capability, latency and context limits (DeepSeek R1 (Reasoner): 64K, o3-mini: 200K) may justify the premium for your task.
FAQ
On input, DeepSeek R1 (Reasoner) is cheaper ($0.55/M vs $1.10/M). On output, DeepSeek R1 (Reasoner) is cheaper ($2.19/M vs $4.40/M). The same model is cheaper on both sides, so it wins at every mix.
A chat-style request (4,000 input + 800 output tokens, 50% cached) costs $0.003132 on DeepSeek R1 (Reasoner) and $0.00682 on o3-mini — DeepSeek R1 (Reasoner) is 2.2× cheaper for that workload.
DeepSeek R1 (Reasoner) supports 64,000 tokens (8K max output); o3-mini supports 200,000 (100K max output). o3-mini fits 3.1× more context, which matters for long documents and agents.