Next-generation AI infrastructure

Turn every unit of compute into intelligence

Model marketplace

Explore leading open-source and proprietary models. Deploy and start calling in minutes.

DeepSeek-V4.1-Flash

DeepSeek-V4.1-Flash is the smallest model in DeepSeek's new architecture family, with native multimodal visual understanding. It is a 552B-parameter MoE model built on the new Causal-Encoder-Decoder architecture, with asymmetric input and output: only 8B parameters are activated for input and 16B for output, at a cost well below other known models of the same size. The new architecture sharply compresses the KV cache, cutting HBM requirements to 1/4 and SSD requirements to 1/8 of the previous generation, which lowers the cost of agent workloads where cache-hit charges make up a large share. The model supports a 1M context window and up to 384K output tokens; thinking mode is on by default, with three effort levels (low, high and max) to match task complexity.

Try now

DeepSeek-V4-FlashLive

1M ContextRAGCode generationContext caching

CHAT

Cache/1M

$0.003

Input/1M

$0.15

Output/1M

$0.6

Off-peak

View model →API access →

DeepSeek-V4-ProLive

1M ContextMathCode generationDocument analysis

CHAT

Cache/1M

$0.022

Input/1M

$0.11

Output/1M

$1.98

Off-peak

View model →API access →

GLM-5.2Live

1M ContextReasoningCode generationAgentic tasks

CHAT

Cache/1M

Input/1M

Output/1M

$28

View model →API access →

GLM-5.3Live

1M ContextCode generationAgentic tasksReasoning

CHAT

Cache/1M

$0.26

Input/1M

$1.4

Output/1M

$4.4

View model →API access →

1 / 2

View all models

Free Trial

Most models offer free trials covering text generation, image understanding, speech synthesis, and more. Sign up to get started.

Try Models

Easy accessYour first model

Access leading open-source and proprietary models through one API.

Get API Key

first_call.py

1from openai import OpenAI2client = OpenAI(base_url="https://api.tokenfab.com/v1", api_key="your-api-key")3resp = client.chat.completions.create(model="deepseek-v4-flash",4    messages=[{"role": "user", "content": "hello, TokenFab"}])5print(resp.choices[0].message.content)