DeepSeek: DeepSeek V3.1
by DeepSeekDeepSeek: DeepSeek V3.1 is a large language model from DeepSeek. It costs $0.250 per million input tokens and $0.950 per million output tokens. Its context window is 164K tokens.
- Input / 1M tokens
- $0.250
- Output / 1M tokens
- $0.950
- Cached input / 1M
- $0.130
- Context window
- 164K
On repeated prefixes
tokens
Who serves it cheapest
8 hosts serve DeepSeek V3.1. Same weights, same API - the price difference is pure margin and routing.
| Provider | Input / 1M | Output / 1M | Context | Throughput | Uptime 24h |
|---|---|---|---|---|---|
| DeepInfraCheapestfp4 | $0.250 | $0.950 | 164K | - | 100.0% |
| SiliconFlowfp8 | $0.270 | $1.00 | 164K | - | 96.0% |
| Novitafp8 | $0.270 | $1.00 | 131K | - | 99.9% |
| AtlasCloudfp8 | $0.300 | $0.950 | 131K | - | 98.4% |
| CoreWeavefp8 | $0.550 | $1.65 | 161K | - | 99.8% |
| SambaNovafp8 | $0.650 | $1.50 | 131K | - | 97.7% |
| Mara | $0.600 | $1.70 | 131K | - | 94.7% |
| $0.600 | $1.70 | 164K | - | 0.0% |
The spread between DeepInfra and Google is 2.1× for identical weights. Quantization and context limits differ, so check both columns before switching.
About DeepSeek V3.1
DeepSeek-V3.1 is a large hybrid reasoning model (671B parameters, 37B active) that supports both thinking and non-thinking modes via prompt templates. It extends the DeepSeek-V3 base with a two-phase long-context...
Specifications
| Model ID | deepseek/deepseek-chat-v3.1 |
|---|---|
| Provider | DeepSeek |
| Context window | 164K tokens |
| Max output | 33K tokens |
| Input modalities | text |
| Output modalities | text |
| Knowledge cutoff | 2025-03-31 |
| Open weights | Yes - deepseek-ai/DeepSeek-V3.1 |
| Released | August 21, 2025 |
Cheaper alternatives
Models that cost less than DeepSeek V3.1 while keeping at least half its context window and every input modality it supports.
Frequently asked
How much does DeepSeek: DeepSeek V3.1 cost?
$0.250 per million input tokens and $0.950 per million output tokens. Cached input reads cost $0.130 per million tokens.
What is the context window of DeepSeek: DeepSeek V3.1?
164K tokens, with up to 33K tokens of output per request.
Which provider serves DeepSeek: DeepSeek V3.1 cheapest?
DeepInfra at $0.250 per million input tokens - 2.1× cheaper than Google, the most expensive of the 8 hosts serving it.
Confirm against the source: DeepSeek official pricing.