Deepseek V3.1 API Pricing
Deepseek V3.1 costs $0.56 per 1M input tokens and $1.67 per 1M output tokens on DeepSeek's API. Cached input is billed at $0.56 per 1M tokens. Prices are provider list prices in USD, refreshed as sources update; see the Deepseek V3.1 model profile for capability scores.
Price List
| Rate | Price (USD) | Notes |
|---|---|---|
| Input | $0.56 | per 1M tokens |
| Output | $1.67 | per 1M tokens |
| Cached input (read) | $0.56 | per 1M tokens |
| Blended input + output | $2.23 | 1M in + 1M out at list price |
What a Workload Costs
Computed straight from the list prices above — token counts are illustrative workload sizes, not measurements of Deepseek V3.1.
| Workload | Cost | Tokens |
|---|---|---|
| Short chat turn | $0.00139 | 1K in / 500 out |
| Summarize a long document | $0.059 | 100K in / 2K out |
| Agentic coding session | $0.447 | 500K in / 100K out |
| 1M input + 1M output tokens | $2.23 | 1M in / 1M out |
Effective Cost
In AI IQ's scoring, Deepseek V3.1's list price is adjusted by a token-usage multiplier of 0.548 (measured from real benchmark runs), giving an effective cost of $1.22 per 1M input + output tokens. That places it #28 of 117 models on the effective-cost ranking (lower is cheaper). Some models spend far more tokens than others on the same task, so effective cost compares what a unit of work really costs. Method details are on the methodology page; see all models on the cost charts and the falling cost of intelligence over time.
Cheaper Alternatives at Similar Capability
Models with a lower effective cost that score within a few IQ points of Deepseek V3.1 (IQ 95), or better.
| Model | Provider | IQ | Effective Cost / 1M |
|---|---|---|---|
| Agnes 2.5 Pro Beta | World (Other) | 111 | $1.22 |
| mimo-v2.5 | Xiaomi | 109 | $0.9553 |
| Inkling-Small | US (Other) | 106 | $1.01 |
| mimo-v2-flash | Xiaomi | 105 | $0.6716 |
| gpt-oss-120b | OpenAI | 105 | $0.7632 |
| deepseek-v3.2 | DeepSeek | 103 | $0.9282 |