Skip to content
Dashboard

DeepSeek V3.1

DeepSeek V3.1 is DeepSeek's August 21, 2025 model update introducing hybrid inference with selectable thinking and non-thinking modes in one endpoint. It strengthens tool use and multi-step agent capabilities over DeepSeek-V3.

Implicit CachingReasoningTool Use
index.ts
import { streamText } from 'ai'
const result = streamText({
model: 'deepseek/deepseek-v3.1',
prompt: 'Why is the sky blue?'
})

More models by DeepSeek

Model
Context
Latency
Throughput
Input
Output
Cache
Web Search
Per Query
Capabilities
Providers
ZDR
No Training
Release Date
1M
1.1s
137tps
$0.14/M$0.28/M
Read:$0.0/M
Write:
+1
azure logo
deepinfra logo
deepseek logo
+2
04/23/2026
1M
0.6s
162tps
$1.74/M$0.43/M
$3.48/M$0.87/M
Read:$0.0/M
Write:
+1
azure logo
baseten logo
deepinfra logo
+4
04/23/2026
164K
0.7s
88tps
$0.28/M$0.42/M
Read:$0.03/M
Write:
bedrock logo
deepinfra logo
deepseek logo
+1
12/01/2025
164K
0.9s
53tps
$0.26/M$0.38/M
Read:$0.13/M
Write:
+1
bedrock logo
deepinfra logo
fireworks logo
+1
12/01/2025
131K
2.3s
27tps
$0.27/M$1.00/M
Read:$0.14/M
Write:
+1
novita logo
09/22/2025
160K
0.4s
158tps
$1.35/M$5.40/M
Read:$0.35/M
Write:
+1
bedrock logo
deepinfra logo
01/20/2025